Articles - GTM

▶ Watch the original lecture

Go to Market: Case Studies and Practitioner Deep Dives

These are the full tellings of the guest case studies and practitioner methodologies that bring the frameworks to life.

Launch Secrets (Jeff, Bright Cove)

I have been part of many product and company launches, and it is an incredibly powerful way of driving value around your business.

Brightcove 4. This is the actual result of a launch we did back in 2009, right after I joined Bright Cove as CMO. The headline appeared on the homepage of TechCrunch, written by Eric Schonfeld, one of their lead writers, back when TechCrunch was the place you wanted coverage. It was massively tweeted, one of literally hundreds of articles that ran that day. What is fascinating is how it came together. I asked engineering what we were shipping that I could go make noise about, and none of it in isolation was that exciting. But in a software-as-a-service model we release things every week, so if we bucketed everything released over a period into specific themes, developed messaging, and put it together, we could launch it as a thing. This launch was made from whole cloth: not a big annual release, just the normal delivery cycle bundled up into a significant launch. Eric was a little onto us; his article asked why Bright Cove holds all this good stuff back until it can package it in a numbered release, since as a web service it could upgrade on a rolling basis, then answered his own question by noting it highlights the changes and shows why Bright Cove is the leading platform for professional use. We got massive coverage, kept our leadership position, changed nothing on the engineering side, and went on to do Brightcove 5.

Leveraging someone else’s noise: the iPad. We had been working on HTML5 video players for a while because we knew Flash was going down the tubes. When the iPad was announced, we got panicked inbound calls from customers worried their video would not play. We had a good story: it works with our new HTML5 players. Realizing the market was hot and ripe, right before the iPad’s commercial availability we went out with the story that Bright Cove is iPad-enabling all of the web’s video. It sounds like a marketing pitch, and it was; this is marketing. But the media was in a frenzy about the iPad, and a conflict story (“the iPad is coming, but unless something happens, content across the web will not work on it”) was compelling. I ended up on NPR talking about customer demand to get content onto the iPad. That does not happen because of my good looks; it happens because the timing was right and we had the right story. It is basic guerrilla marketing: feed off the noise someone else is creating and insert yourself into it. We did the same with Chromecast: Google shipped its open developer SDK, we had been working with them and scrambled an announcement of support, wrote a detailed article, pitched it to recode (Kara Swisher’s venture), and they reposted our CEO’s blog post verbatim as an original article because they were looking to make sense of Chromecast.

Country launches: Dubai. It is not just products that you launch. Bright Cove is a public hundred-million-plus company, yet last week I was in Dubai launching our Dubai office, because we take nothing for granted. We have a playbook. First, develop a target list of media companies in the country and approach them directly with sales to close a handful of referenceable customers. Then secure local real estate and legal infrastructure, hire the team, and get a local PR agency with real local contacts. Then book a press conference with the key trade press: in Dubai we booked the top-floor conference room of the Burj Al Arab, the seven-star hotel, invited all the top press, did great one-on-ones, and presented material specific to their market (research on internet and smartphone adoption and video consumption, localized to their market) with our own reference customers there. Previously I launched Seoul, Singapore, Sydney, Barcelona, and Paris the same way. We have a saying at Bright Cove: we should always act and feel ten feet tall even if we are only six feet tall, and we accomplish that by executing launches effectively.

On making messaging relevant. The best messaging connects with real pain that someone important is feeling. We start our campaign plans from people and their pain, and we work messaging next. That is what makes it relevant.

Brand: Demandware, a Living Case Study (James Driscoll)

This is an ongoing, living case study; the execution is not yet live, so you are getting a sneak peek. When you are small, a couple of people in marketing, you are the brand and make choices instinctively. But as we grew onto a much larger platform with a much larger microphone, it was time to institutionalize the brand: what attributes do we want to build equity around as we execute in many markets? We analyzed our customers and the broader market to find the attributes we are known for and the challenges we face.

Fundamentally we are trying to change the criteria in the buying cycle. Demandware sells an e-commerce platform into a market that already had a software space, but we deliver it as an on-demand service, so our marketing goal is to change how the market evaluates us: from cost to advantage, from managing software to growing the business, from mid-sized companies to high-growth brands. When we mapped the market, most players sat in the blue boxes talking about e-commerce or technology as features or benefits. The great opportunity was in the green squares, the intersection of the rewards individual buyers want (their aspirations and goals) and the big objectives of large retail organizations (brand and marketing).

One of the great opportunities in enterprise software now is to talk about the emotional side. We use iPhones because they are cool; could enterprise software be something individual buyers aspire to, something they mention with pride at a cocktail party? So we position around buyers’ ambitions. Picture a kitten looking in a mirror and seeing a lion: we want buyers to see reflected back what they think of themselves based on the selections they make. There is a long-standing axiom in marketing: people buy on emotion but justify with fact. So while finance justifies with better economics and the CEO with growth, what they really want is to feel something (IT wants a stronger seat at the table, the e-commerce person wants to advance toward the CEO chair, the CEO wants to be a leader to watch), and our branding attaches to those aspirations.

Our brand promise rests on the concept of potential: we are the means by which they achieve their dreams, the canvas on which they paint their masterpiece. It stands on four pillars: innovation (help them innovate faster and better), simplicity (it is easy, not hard), partnership (in it for the long haul), and performance (support their ongoing success). In the rollout you will see green (innovation and growth) and gray (steel and strength), the customer’s brand imagery featured heavily because it is about them, and a repetitive language formula: the customer and their brand come first, their comparative verbs next (we help them do it better), “with” as a statement of partnership, and we come last. The home page will say, “Great brands have great ideas. We help make them possible.” These brands are willing to let us use their brand, which is rare in software, and that rings loudly as a statement of partnership. The lesson: do not stop at features and benefits. The ultimate measure is the reward, the emotional quotient, because that is what creates the connection.

Notably, the Demandware promise (simplicity, empowerment) was made very early on, eight years ago; it barely changed. We just reached the point where we can back it with a customer voice.

Brand: The Brand Essence (Adam Barry)

Think of a brand you love and why it is great. Audience answers: Dell, because it was cheap and affordable; a product for its sameness; Moleskine, for the perfectionism in every detail of quality and paper size. A strong brand drives sales, opens markets, and drives top-line revenue and profitability.

Normally when you think brand you think of an expression of it: the Nike swoosh, the BMW grille, the Nordstrom service, the Harley-Davidson sound. But those start from a deeper core idea, the brand essence, and everything else (name, logo, website graphics, the voicemail voice, how the site services customers) is an expression of that essence. The essence starts with you, the founder, and the culture; if the brand is inconsistent with the culture, it will break as you grow.

A brand essence has four components: a vision (how the world is changing and how your product reacts to it; Nike tackled the amateur-athlete and personal-health movement of the 70s); a promise (Nike: perform at your very best, run as fast and jump as high as you can); attributes at three levels (the spike or one simple thing, such as Zipcar’s freedom, Nike’s performance, Puma’s style; three or four differentiating attributes, such as ColdFusion’s speed; and cost-of-entry attributes such as safety in cars or scalability and reliability in databases); and personality and style (if the brand were a person, what personality and what would they wear? Chanel classy, Victoria’s Secret sexy, Oracle stark, Apple elegant; Steve Jobs was literally a manifestation of Apple).

Positioning: Spotfire (Mark Lorey)

Positioning statements may look academic in the audience, but they are far from it. Go through this early and often, particularly when forming, launching, or pivoting, because the ultimate path you are on is very unlikely to be the one you start on. Every word in the framework should mean something, and if you read it from a competitor’s standpoint, the whole thing should fall apart.

Spotfire was a private company acquired by Tibco in 2007. Around 2005 we went through a significant pivot. We had been targeting mostly scientific users in data visualization, selling to small teams in corners of organizations. We wanted to pivot to the executive ranks and the business-intelligence space, which had bigger budgets, but we were virtually unknown there, and anyone who knew the brand thought of us as scientific and technical. As founders and early people often do, we loved the features, so we gravitated to what the product does rather than what it does for the buyer. Going up against Cognos, Business Objects, and Hyperion, well-funded with big marketing budgets, we were never going to win by claiming to be another BI tool. So we took a different angle: the value of decisions. Spotfire helps people make better, smarter decisions in every corner of the company, and it is fun and easy to use. We fought through the positioning statement as a management team; it was ugly. We would not move forward until we had alignment on the management team, and then we went out in concentric orbits; the CEO would get in your face if you could not recite the framework. There was definitely a rebranding effort, and then campaigns and programs fell in line because everyone was singing from the same hymnal.

Positioning, live: Fair Setup (audience workshop)

An entrepreneur in the audience volunteered that his startup does compensation and has no direct competitors. But others are solving the compensation problem, and customers would have to find money from an existing budget for an incremental improvement, and it is a cost-saving proposition. So the challenge is the one I hear over and over: you feel unique, but you are in a category with a finite pool of dollars, fighting it out with everybody. On the spot he articulated a positioning statement: an app for CEOs and upper management at mid-size and large enterprises who struggle to drive employee behavior; the approach integrates into existing meetings to calculate each employee’s impact with a light, minimal-overhead weekly method perceived as fair and connectable to compensation, unlike the performance-evaluation approaches everybody hates; a new approach to performance management and compensation that all participants find fair, improving engagement and performance. Impressive on the spot. You will likely be doing that for the next five to eight years until finally people say, “Fair Setup, I know those guys, best compensation system on the planet.” It took Coke a lot of money to get to where a single word conveys everything.

Segmentation and Whole Product: Demandware (James Driscoll and Zach)

We faced a very large market: e-commerce, thousands of entities in North America alone, and we were about twenty-two people. Our guiding principle was “get small, get big, or get out”: pick an area small and discrete enough to win in and get really big within that niche, or move to another niche. Find a market small enough to be actionable, where everybody knows everybody (for reference selling), but large enough that winning it is meaningful, where we could deliver with distinction, and where winning would open up other niches.

We looked at the market in many facets, not just revenue size and industry but hard-to-pin brand attributes like style and growth aspirations, and positioned around high-growth retailers and brands: within retail, the highest and fastest-growing. We wired specific account sets directly into our databases so sales and marketing were aligned around highly referenceable accounts (apparel, footwear, cosmetics, luxury, home, lifestyle above certain revenue sizes). It started with one brand; that brand would tell friends they made a great choice with a nice startup doing wonderful things, and word spread. It worked city by city: win a brand in a city, and once that customer signed on and believed, the whole city opened up. Segmentation built into positioning, wired through operations, permeated everything we did.

The link program (whole product). E-commerce is a complex end-to-end solution: platform, shipping, tracking, consumer engagement, ratings and reviews, back-end fulfillment. We could not deliver everything anyone envisages, and a recurring brake in the sales cycle was that Demandware was a piece of the puzzle, not the whole puzzle, so other technologies and vendors got pulled in, encumbering the cycle and driving up integration cost, which stalled deals. Rather than us doing point-to-point integration with every partner, we turned the model on its head and asked partners to integrate with us. Because we are an on-demand platform, the technology never really changes, so an integration a partner does once is repeatable across every client. Now lots of link partners pre-build integrations, talk about us, and give away those integration assets at no cost to our clients, because it helps them get additional clients. It hit multiple dimensions: it increased reach (multiplier), took out cost and structure (lever), and because these are revenue-generating technologies added quickly under our shared-success model, time to market and time to return are faster. Competitors who sell enterprise software have a very difficult time matching this repeatability and margin. We launched the program in April 2010; the early curve was shallow, then once we proved value it accelerated, and we now have a little over one hundred fifteen partners and are more selective.

On the partner chicken-and-egg: initially partners will not invest time in you until you are proven. Through our segmentation and initial go-to-markets, once we brought on a couple of marquee clients, we had something to take to the partner community. Then link showed them a more successful, repeatable business on top of Demandware, giving them added incentive to work with us.

Demandware later went public under the symbol DWRE.

Targeting and Brand, from Customers: PatientPing (Jay)

We are building a communication network. When any patient goes from one facility to another, everyone in that network is notified. If I am a primary-care doctor and my patient goes to an ER, I get a ping. Hence the name PatientPing.

Before this I worked at the government at CMS on Obamacare (everything but healthcare.gov), on accountable care organizations, a new concept putting financial risk into providers’ hands so they take accountability for outcomes and keep some of the savings. A big problem there is tracking patients when they go outside your own network, called leakage. Partners, with Mass General and the Brigham, does a good job inside its own facilities but struggles when patients go outside. We tried to fix it at the government by getting states to build health information exchanges, but nobody was really doing it, so we saw an opportunity to build a network from the ground up in a radically different way.

We segmented starting with health systems, then accountable care organizations. I was able to call all thirty-two Pioneer ACOs, the bleeding edge, and ask: is this a big problem? Yes. Is this the state of the art? Pretty much. Does it suck? Yes, terrible. What if we did it this way? Oh, that is interesting, that would be great. So our brand and positioning were to be slightly different from the state of the art and address the pain points of a disjointed, imperfect system. The buyer is the population-health management team; at Partners that is run by Tim Ferris, so he is our buyer there; at Steward Healthcare there is a population-health group with someone in charge. Every organization has a team managing a panel of patients on whom they are not taking cost accountability, so the need is consistent. There are opportunities to reduce total cost by coordinating care when patients receive it outside the network. If they do a good job with sixty of a hundred patients, there are forty more where there is opportunity but no visibility, just darkness, and they want light. We are trying to deliver ten times the value we charge for.

On brand from experience: our go-to-market launch was a cliché (all-nighter, ramen noodles, things breaking the night before), and then customers started saying, “This is really great, very simple, I get it, easy to use,” and that was when we realized that is our brand, though we had not consciously set it. We also learned our values by contrast: a candidate compromised the integrity and security of our patient data, and the contempt we felt told us that data integrity is core to our brand. We have since written down our values: simplicity, quality, clarity, value, and integrity. And there is the flip from “you” to a company: pre-market it is your story that you tell employees, customers, and investors, but once you are out to market it becomes their story. When a big customer’s person said, “Did you hear about the thing Jay’s doing?” that is not good for us; we want it to be “Did you hear what PatientPing is?” Flipping from the founder brand to a consistently executing company brand is extremely hard work. And start simple: the only way people capture and re-share your story pre-market is if it is very clean and simple, so we spent a lot of energy making our solution easy to grasp.

Segmentation, a B2C Journey: Splash Score, formerly Aipia (Lyle Stevens)

We are an influencer activation engine for consumer brands, a two-sided marketplace with a large B2C component, which makes go-to-market very complicated because you have to pick one side or the other. We identify and activate influential consumers to give brands an army of advocates. I will tell the story through “startup vitamins.”

We were inspired by the internet troll meme. My two younger brothers, in middle and high school, were competing over who was the better troll, tracking it in a spreadsheet, and as an engineer I thought, why not build something to track this? That led to the epiphany: what if we could identify not trolls but influencers online? First vitamin: holy crap, we might have something, let’s just do it, because there is a bigger risk in not doing it.

Where do you start: research competition (like Klout), define the product, name the company, define the customer, or define channels? Most people vote for defining the target customer first, and you cannot pick channels without knowing the customer. We did the wrong thing: we picked our name and value first. We took idea and epiphany, respelled them, and became Aipia, with a vision to be “PageRank for people,” and three core values (customer happiness, continuous improvement, consistent results). We started building an algorithm to measure your “ripple” on your “pond” of friends and content, the ratio being your “splash,” which was pretty stupid in hindsight, but the lesson is you need to begin somewhere; begin anywhere, because you learn as you go. We built a stupid-basic MVP inside Facebook (our first channel, because that is where people spent time and we did not want to change behavior), and getting something out beat making it perfect.

Then we picked a target customer. Considerations: ages 13 to 18 (the original inspiration), 18 to 25 (college), 25 to 45 (professionals), or 40 plus. We launched with college students at Northeastern, got about a thousand signups in a week, then dived into the data and realized college students passively observe on social networks; it was their mothers, aunts, and older sisters who were extremely active. So we switched to women 25 to 40, which became our minimum viable segment. We literally built a Twitter list of 5,000 women with “SAHM” (stay-at-home mom) in their profile, tweeted at them, and brute-forced a thousand signups. That became our first persona, Household Helen (we pictured Helen from The Incredibles): balancing a lot, cost-conscious, runs the family, powerful. But she did not care about the score, did not know she was influential, and asked what to do with it. As you quoted from Steve Jobs, if a user has a problem, we have a problem.

So, “innovate or die.” We watched Klout, which tried to become the credit score for influence but floated along until a partnership with Spotify’s US launch made them an overnight success. The aha: people do not care about being influential, they care about getting stuff for being influential, which Klout never dug into deeply and we did. We rebranded to “make a splash on Facebook, score free rewards.” Then we chose which brands to pursue: startups, SMBs, local, geolocal, regional, or large US brands? We went after large national and global top-200 brands, but not the number-one player, rather the number two or three trying to catch number one (like Reebok chasing Nike), because you can always go backward to local brands, but once you do local and SMB the big guys will not take you seriously. Lesson: think bigger.

To reach them we did tons of interviews and developed two personas: Marketing Molly (a digital director at a large consumer brand trying to reach consumers in social, whose number-one channel is Facebook, which had just tweaked Edge Rank so organic reach was plummeting, making us more needed) and Agency Amy (works for Molly, cares about looking good to Molly and doing simple, easy things, a bit more of a digital native). When we said we could identify their influencers, they said they could already do that; what they could not do was activate them to drive real value. Everyone out there was good at either identifying influencers (Klout) or activating people who were not influential (contest apps), so we fit the white space of identifying and activating influencers, dropped Aipia entirely, and became just Splash Score. We help brands deliver engaging content to influential people on social media whom their friends already trust, because you trust a friend’s post, not a display ad. Our value prop today: twenty times more clicks, likes, and leads than any other social ad. The lesson: start somewhere, you will learn, you will fail, and you will repeat.

Reflecting: we would use one brand, not two (building Aipia for brands and Splash Score as the product simultaneously was hard and expensive; as a startup do few things well). Competitive analysis was very important, because Klout’s mistakes were our free testing; we did not have to learn those lessons the hard way. We should have talked to who we thought our customer was before building anything, and talked to brands sooner. And on the dual-sided marketplace: we essentially flipped a coin to pick influencers first, but the better guiding light is to pick the side with the greatest pain point where you can meet a real need first, because that lets you pull the other side.

Startup Secret: With a dual-sided value proposition, pick the side with the greatest pain point and real need first; it will pull the other side.

Inbound and Belief-Based Selling: Unidesk (Tom Rose, Andrew, Brian)

What Unidesk does. You are used to accessing your mobile apps anytime, anywhere; Unidesk gives that same access to Windows virtual desktops. Thousands of Windows applications are not going to the cloud anytime soon, and customers want them on iPads, home PCs, work PCs, thin or zero clients. Unidesk makes it “stupid easy” to provision those desktops on a virtual server, deliver applications, patch and update them, and fix them if an end user breaks them. We are pioneering the new standard of desktop management for virtual desktops, replacing old PC-management packages.

Selling to belief. Simon Sinek says to find customers who believe what you believe and to talk about your beliefs, not what your product does. His examples: Martin Luther King, who talked about his dream, and Apple, whose “1984” sea-change commercial sold individuality (a personal computer to divorce yourself from being the same as everyone on the mainframe), and whose later “what’s your verse?” campaign spends two minutes on making your mark on the world before the iPad appears at the very end. That is what creates zealots.

Unidesk believes managing desktops should be really easy; you should not spend your life patching PCs, delivering applications, and fixing broken desktops. So who shares that belief? A Wall Street bank with 2,000 people in IT celebrates complexity; its IT managers want to be called when a desktop breaks; they do not share our beliefs. But the VP of IT at Needham Bank with two IT people (responsible for servers, storage, phone systems) absolutely shares them, as does the director of IT at Mass Maritime College and the head of IT at Bernstein Shur law firm. So our target is organizations under 5,000 people, which tend to have lean IT: state and local government, K-12, higher education, mid-sized healthcare. That is our tribe. The next level down, our segment: we are a virtual-desktop management play, so we go where the virtual desktops already are, customers who have bought VMware virtual infrastructure and chosen VMware View or Citrix XenDesktop. The actors: the director of IT, the Windows admin, the help desk, the people whose lives we make better, who get home earlier for dinner and become ardent advocates. We had the VP of IT from Needham Bank come in, and asked a law-firm customer to dial in for fifteen minutes to explain why they chose Unidesk; no problem.

Social media (Andrew). A common misuse is treating social media as an island, hiring an intern to “go do social media.” Social media has to be part of your overall content strategy and, one step further, part of your company culture. What impressed me joining Unidesk was that all our executives are on Twitter constantly (Brian, Tom, our CEO, our CTO and founder), putting a personal touch behind the brand, which goes a long way for an early-stage startup building a tribe. Qualify which networks to spend time on: we sell to IT people who are on Facebook but do not want to hear from Unidesk there or share our content, because their friends are not in IT. If you are B2C with a visual product, Pinterest may be right; a travel company seeing an uptick in China should be on Baidu (the Google of China); five years ago building a social network for foreign-currency traders, our big markets were the Philippines and Indonesia where Facebook was not big, so Hi5 was a big channel. Unidesk, being B2B, uses YouTube (the second-largest search engine, and IT people love finding videos of solutions), Twitter (not great for lead gen but great for cultivating the tribe; about fifty customers are so engaged they reply on our feed before we do), and LinkedIn groups (inserting ourselves where people already are).

Three examples. Sanjay Poonen, head of end-user computing at VMware, is the main person we need to influence; we have a “frenemy” relationship (we compete on one level, but every Unidesk license requires a VMware license). Every year at VMworld we staff our booth with unpaid customers who happily give demos; a higher-ed customer talks to a higher-ed prospect. Our CEO Don Bulens (Grouchy CEO on Twitter) talked to Sanjay, Sanjay tweeted it, and within minutes five customers replied, two in the booth and three from around the world (Pete Sears in New Zealand, two in the UK). Nineteen retweets and five favorites made it one of Sanjay’s most popular tweets. The lesson we learned: instead of positioning against VMware’s flaws to make our light shine brighter, we shifted to selling along the shared belief of joint success for common customers, which is where we started to grow and scale, and it motivated VMware reps to treat us differently. Second, someone asked in the VMware community forums for a Unidesk-and-VMware review; we pinged customers and within hours had four responses and eight replies, an unbiased review that now lives on the forums and in Google and arms our sales team. One respondent was Trek Bikes, whose detailed reply we repurposed as a case study in twenty-four hours (added their logo, kept the content). Your case studies happen in the strangest places; our customer webinars, forty-three of them, are living case studies that a marketing manager transcribes for quotes, with fast approval because the customer already said it. Third, accelerating the sales cycle: Kellen Dam tweeted he was ripping out Mirage (a competing VMware product) and checking out Unidesk; within hours customers replied “Unidesk is king, good choice,” and our chief solution architect Ron Oglesby offered help. That day Kellen tweeted, “Judging by my Twitter feed, I’m making the right decision,” eight days later after testing said “your product really is stupid easy,” and five days after that chose Unidesk, a fifteen-day sales cycle against our average of seventy-five days.

Support as the most overlooked marketing tool (Brian). Things were not always perfect; our first product had reliability challenges we had to muscle through. One of the most important early decisions was to over-invest in incredibly capable customer-support people, because the best way to disrupt this market was the voice of the customer. It fostered incredible loyalty; we can call any of our 600 customers and ask them to dial into a pitch and they will, from home, because they love to talk about their experience. The beacon through the stressful times was that beacon: customers kept telling us we were on the right path and if we pulled it off it would be life-changing. Your first version may be hard to implement, lack features, or have reliability issues; all three can be overcome with great support, so over-invest in support to buy your engineering team time. We hired support people who understood the entire stack, so when a customer called, we owned the problem even if it had nothing to do with Unidesk (much of the scaling trouble was storage, network, and performance intricacies), and we never left them. We suffered with the product issue for over two years.

Channel (Brian). A channel cares about your beliefs and, ultimately, what their customers care about, so you must be completely aligned and qualify carefully. Think of the desktop-virtualization stack as six decisions a CIO makes to deploy VDI, from client down to storage, with Unidesk layered in the middle; small businesses lean on a trusted advisor to consolidate those decisions. So our early focus was to identify the best VMware channel partners to tap pipeline they already had in flight. When defining your channel, ask what products or services are bought before, after, or alongside your solution and how those are sold today; chances are you do not have to blaze a new trail. Build solution briefs so customers can see you working with IBM, Cisco UCS, Samsung zero clients, and reference a customer who chose both, with a landing page, to short-circuit the process. It is about leverage: we close one in four of our online demos and fifty percent of our pilots, and create a ten-times drag on infrastructure for solution partners, which motivates them. The channel helps you skip gears and “hijack deals in flight,” reaching people already beyond the understanding phase. There are actors in the channel too: a technical gatekeeper decides which new technologies get taken on, so we armed our team to win the technical mind-share first, and the sales team fell in place. We got turned down a few times by gatekeepers who did not believe, but won many who did. And do not run to the channel before proving you solve a problem for the end customer; win the customer first, then ask who they buy from, then tell that partner “I just sold your customer, would you like to be involved?” It is amazing how much credibility that gives. We are now in a hybrid model, touching both customer and channel, but all of it architected through the eyes of the customer.

Removing brakes and engaging the clutch: marketing and sales automation (Tom, Brian, Andrew). A brake in our business was a very manual, labor-intensive lead-follow-up process: lots of voicemails, people not ready, no budget, no authority. So over six months, led largely by our CTO and founder Chris Midgley, we integrated HubSpot as the marketing engine and Salesforce as the sales engine, surrounded by other tools, to automate identifying VDI projects, nurturing them until ready, and then efficiently reaching them. We use calling technology called Connect and Sell: a typical rep might connect with six or seven customers a day, and Connect and Sell increases that three or four times. GoToWebinar registrations flow automatically through HubSpot into Salesforce; Wistia captures video views; scores rise as prospects consume assets, raising a flag for Brian’s team. Our selling workflow has clear stages (lead, marketing qualified raw lead, marketing qualified lead, sales qualified opportunity, POC, won), with assets and automation rules at every stage, and once won it loops back for repeat and upsell. The effort is in setting up workflow and rules; once set, it is a scalable engine, and you just pour more into the top. Marketing and sales is math: monitor the stats and turn the dials; do not just throw money and hope. A live Salesforce report tracks every stage and the conversion rates so we can find where we lose prospects. Stay close to customers, document their buying process, automate where you can, measure and score everything, let the metrics drive behavior, and focus on making your customers unbelievably successful, and magic will happen. Since Unidesk was last here, they have doubled the business year on year two years in a row, and they have always underspent versus plan because of how clearly they measure everything.

Guerrilla marketing across the stages (Tom, Part 4 telling). Unidesk is my fifth startup, so I am used to a limited budget, and we applied guerrilla marketing and web marketing from inception. In the requirements phase, four years ago with no product, we used low-cost Survey Monkey surveys, capturing our CTO Chris Midgley’s innovation reputation in emails to an acquired list that said, “I have nothing to sell you, I just want your ideas on how to transform desktop management.” About seventy-five people responded, writing pages about their challenges. In pre-launch, we started blogging and building our long-tail SEO and keywords about a year and a half before we had product, which drove inbound activity and made it much easier to get beta customers. At launch, the press is gone; you are the media, and so are the influential bloggers, so we gave the key desktop-virtualization blogger an exclusive to unveil us. To build pipeline, we built crisp product-demo videos (Apple-esque, two-minute shorts up to a fifteen-minute max) featuring compelling personalities like Chris Midgley, and put them on YouTube, LinkedIn, Twitter, and the blog. For customer acquisition, we run webinars every three to four weeks featuring our passionate customers as the stars, focused on verticals (one today was healthcare, mostly hospitals and clinicians), which drives references at very low cost per lead. And in production, we opened Drupal-based forums (the platform by Acquia) with search across blog, knowledge base, support, and forums; about a week after opening we already had 215 posts and nearly 100 replies, showing prospects a dynamic, growing community. Through it all, measure everything and refine the funnel.

Channel, the science (Brian, Part 4 telling). Two reasons we went to a channel model: five decisions a customer makes before or as they purchase Unidesk, one of which is VMware infrastructure, and a way to target and market to those customers. So we identified the top VMware partners selling into our segments (state and local government, higher education, specifically faculty, staff, and knowledge workers), refined to partners who would take a chance on a disruptive technology, choosing regional boutique early adopters over national laggards with long onboarding. Then refine and focus further: identify the individual the sales team listens to, typically the trusted technical architect, and target that person. Once we gained the lead technical architect’s endorsement, we had permission to engage the sellers, whose attention we could not get otherwise, with the goal of intercepting partner-led opportunities and interjecting Unidesk. Our bold statement to the channel: “We will not let your customers fail. We will do all the work, all the selling, all the support through the free pilot, make them amazingly successful, and they will call you to tell you.” If you get channel time when you have no customer evidence, you probably engaged the wrong partner. When scaling to larger partners, we invite one of our customers (not even theirs) to the recruitment call to tell the story, and for whatever reason the partner listens. We are maniacal about measurement: no easy way to score the channel, so you end up in a spreadsheet, categorizing partners as A (strategic, 60 to 70 percent of revenue) and B, and going a level deeper to ask what percent of mind-share you have in a partner (is all success from one office or set of reps?). The channel is good at field-based marketing but not at social-media marketing or nurturing, so get them to share their database and let you nurture it. Then repeat the cycle.

Inbound Marketing: HubSpot (Mike Volpe)

HubSpot is a software-as-a-service marketing platform. Our differentiation is two things: we focus on and enable inbound, and we are an all-in-one platform (social media, blogging, marketing automation, personalization, lead scoring, analytics all in one place, so it all connects). We focus on the mid-market (companies from 10 to 2,000 employees) yet have 10,000 customers in 56 countries with only two offices, and some very big brands.

We sell to people who believe what we believe: that the way we all buy has fundamentally changed, that people skip TV ads and are on the do-not-call list. People who really believe it have a fast sales cycle and become super-successful customers; people who do not believe it hold us back.

If you are not doing inbound, you are renting the capability of building an audience from someone else: pay Google for AdWords and Google has done the hard work of building the audience and rents it back per click; the same for a newspaper or website ad. The problem with renting is that the moment your budget dries up, the asset is gone, because it was never yours. More than half the leads we generate at HubSpot are from things we did not do this month; one blog article I wrote six years ago still gets me about fifty leads a month, totally free, an asset that keeps paying dividends. To build this, stop thinking like a marketer and think like a media company: be a media mogul, be Oprah for your market. I was talking with a woman selling Montessori toys for young children; she needs to think about the blog, video, or podcast that parents like my wife and me want to consume, related to her business, not about her products. Our competition is broader than other marketing-software vendors; it includes the sites where people go to learn (MarketingSherpa, MarketingLand). We have been at this seven years (I was the fifth employee); we have over 6,000 blog articles, write over 100 a month, and have hundreds of webinars, videos, and presentations. Our blog gets 1.5 million visits a month that we pay nothing for, a huge opt-in email list, hundreds of thousands of Twitter followers, over half a million on Facebook, a LinkedIn group over 100,000, and 56,000 company followers on LinkedIn.

This makes marketing itself a competitive advantage. Being good at buying AdWords adds no inherent value to your company, but having more Twitter followers, links, and Facebook fans than Salesforce.com is an asset you can leverage to acquire customers faster and cheaper, and it is hard to catch up: there is no store where Marc Benioff can buy another 20,000 links into his site, even in Bitcoin; he would have to do the hard work of writing thousands of great articles.

Our funnel: a few million visitors and about 50,000 new opt-in leads a month, from a huge spread of sources (word of mouth typing hubspot.com, the blog, SEO, social media, free tools). Eighty percent of our lead generation is unpaid, the reverse of most companies, and inbound leads are about thirty percent cheaper with double the conversion rate, so sales, the CFO, the CEO, and the board all love them. We still do some paid to fill gaps and experiment, but it should not be your main lever anymore.

Accelerators, brakes, and clutches for us. Accelerators: a free trial (huge, because many competitors do not have one, so our reps proactively offer it and suggest the customer ask competitors for a trial too, which sounds suspicious when they cannot provide one); Marketing Grader; educational content about inbound (we build our own “religion” so more people believe what we believe); and prior product use (a gigantic accelerator, so we try to detect when a customer user changes companies). Brakes: people unwilling to embrace inbound, and conservative buyers who oddly think a company with 10,000 customers and case studies is not a safe play because we are edgy. Clutches we engage to overcome bumps: calls to action on content; heavy reference selling (we launched seven case studies this month and seven last month, and have an internal program so reps can quickly get happy customers on the phone); and third-party validation on B2B review sites like G2 Crowd, TrustRadius, and VentureBeat, where by encouraging happy customers to leave reviews we became number one rated on all three (which does not happen overnight).

Two examples. Marketing Grader, built by our technical co-founder Dharmesh the month I joined: type in your URL and get a score and feedback on your online marketing, free, in a minute or two, fully customized. People ran their own site, then their friend’s (“I got five points higher than you”), then VCs ran all their portfolio companies (“Why’d you get a 42?”), and nice VCs like Michael sent congratulations on an 85. Over four to five million organizations have run their site through it. The key: do not just give free info with no follow-up; make the next step really clear (“Want an easy way to fix all these problems? Click this button”). Second, our blog, which we use to educate (marketing speak for brainwash) people about inbound: over 6,000 articles, hundreds a month. The issue is that many company blogs have no next step, so at the bottom of every article you must add a call to action. When we went from no CTAs to CTAs, we tripled blog leads within a week. Get more sophisticated by targeting the CTA to the article’s content, and by where the person is in your sales process (a customer sees a conference invite; a first-time visitor sees an educational ebook; someone in the middle sees a free-trial offer). Everything links to a well-optimized landing page to gather information.

On mobile apps: the app stores are essentially search engines, so App Store optimization is your SEO; provide value early with a free version or in-app purchases; and use virality within the app and by piggybacking on existing social networks (a lot of early Pinterest growth came from spamming the Facebook feed) to grow, choosing the right networks for your customers.

On staffing content: about fifteen percent of our marketing team is 100 percent dedicated to producing content, and the rest spend some portion of their time. I take dollars from advertising and hire content people instead; my VP of content even cut a deal with the paid-advertising owner to fund a new hire from that budget as long as he drives enough new articles. We use the acronym DARC (digital, analytical, reach, content). The bad news for journalism but good news for marketing is that you can hire journalists into content roles; the tough part is highly technical topics where a journalist may not credibly write for the audience (Unidesk’s IT folks), in which case use customers to create content and have a journalist polish it, or make an industry expert outward-facing.

On how early to start content: never too early, even before you have a product. A great way to think about market validation, and a great asset to walk into a seed or angel meeting with, is: “I do not quite have a product, but I have been blogging about this topic for a year; here are the posts, followers, and traffic, the questions people ask, and a survey of my 10,000 readers about their core problems and what they would pay to solve them.” That is fascinating, as opposed to spending all your time building the product. And if you cannot effectively build an audience to read your blog, you are probably in the wrong market or thinking about something wrong. Doing this early also signals to VCs that you are thinking about the customer first, whereas most people think product first, then customer.

Inbound Marketing, the Thesis: HubSpot (Brian Halligan)

Inbound marketing rests on two observations. First, a radical transformation in how humans live, shop, and learn, a tear in the fabric of the universe. My dad came home at six, had his scotch and water, and opened all his mail and read it; I never open my mail because there is never anything useful. He got seven TV stations (channel 2, 4, 5, 7, 38, 56, and channel 68 with the rabbit ears just right) and talked on the phone a lot. I am on Twitter, Facebook, LinkedIn, and Gmail. Second, the marketing playbook (buy a list, bang people with email, hire young telesales reps to cold-call, spend on AdWords, hire a PR firm to interrupt journalists, do TV and radio) worked my whole career, but its problem is you reach a lot of people who do not want to hear your message. People are sick of being marketed and sold to, and block it out with DVRs, caller ID, ad blockers, spam protection, and priority inboxes. So take everything you learned in marketing class here and throw it away, and rethink marketing to match how humans actually shop and learn. If you start a new company, do it with inbound.

Two things I love about inbound versus outbound. First, success is more about the width of your brain than the width of your wallet: big companies have thick wallets and thin brains, small companies have big brains and thin wallets, so inbound is great for small businesses (and I hear HBS people have big brains). Second, the way it scales. Most venture-backed startups shovel their Sequoia money into Google’s furnace with AdWords and Facebook ads; if you get the math to work, you put in a dollar and get a dollar-ten out, and it is stuck. Inbound is different. If you are CEO of Ford, your balance-sheet assets are factories, inventory, and cash. If you are a marketer or founder, your marketing assets are links into your website, Twitter followers, Facebook fans, keywords you rank for in Google, and pages on your website. You create a piece of content today and it is an asset that lasts and scales essentially forever, pulling in customers; you own it, you are not renting space on Google or Facebook. And people hate outbound (no one likes being called at six o’clock or getting spam or TV ads), whereas inbound content is rich, informative, and engaging, so people fall in love with your brand the way they love Patagonia, Apple, or Whole Foods.

How to do it: turn your website into a modern magnet by creating tons of remarkable content (brilliant blog articles, ebooks, webinars); the better it is, the more retweets, likes, and links it gets, the longer it sustains, and the more leads it pulls. Think of yourself as a production studio like Disney, Fox, or CNN. The internet is a set of pages (websites, the big ones being big sites) connected by links, and the more links you have, the more visitors, authority, and mojo; links are to the internet as dollars are to the economy, and you get links with brilliant content. Your website starts like Cambridge, Massachusetts (zero airports, no real bus stations, a couple of highways) and you want to turn it into New York City: bus stations are Twitter, train stations are Facebook, airports are LinkedIn, and the highways are links from other websites.

My favorite philosopher is Warren Buffett, who tells his CEOs to build a moat around the business, make it wide and cold and put sharks and alligators in it. The modern moat is not a patent or trademark, it is this inbound stuff: how many links, keywords, and fans you have and how they are converting down the funnel, which is really hard to replicate. Take Zappos: if you and I wanted to bury them, we could copy the good-looking website, the funky culture, the inventory, and the supply chain, but we could not replicate Tony Hsieh’s six million Twitter followers, the 500,000 links into their site, and the five million keywords they rank for. That is a nearly insurmountable competitive advantage, and it is what I wish for you. Check out the inbound marketing book, or go to the HubSpot Marketing Grader, put in your URL, and get a grade of one to one hundred.

The Sales Funnel and Micro-Steps: David Skok

Let me cover the funnel in a slightly different way. Take a bunch of suspects, put them through stages, and hope to get a closed deal, then reverse it to expand usage and upsell. In a perfect world you would do it all in one step: put up a website with a small product video, say it costs $9,999, and a Buy Now button. Why does this not work? Ask it of your own product. Reasons the audience gives: you are not targeting a specific-enough segment; even with the right buyer, they see the price before the video and worry about ROI; they do not trust the product will work; there are multiple stakeholders who make decisions; they want to dig deeper for confirmation and there is nothing else to do on the page; they want affirmation and social proof from other customers; they have a fear of failure and of looking like idiots for bringing in something that did not work; and they want to know it will actually solve their problem.

So step out of how you think about the world and into the customer’s brain, and write down every one of these points for your product (will it integrate with Salesforce, will my people be able to work it, is there good ROI, and above all fear of failure). The art of marketing is taking what should have been that one instant-buy step and designing a series of steps that answer each of those questions.

My buying cycle is simpler than Michael’s, with three stages: awareness, consideration, and purchase. Consider the sweater store. You wander in with no intention to buy, and within five seconds a salesperson rushes up and will not leave you alone; it is not enjoyable, because people do not like to be sold to. Now you spill coffee on your sweater before an important meeting, run into a store desperately needing another, and cannot find a salesperson; very irritating. The difference: in the first you are very early in the buying cycle with no awareness of what you want; in the second you are far down the purchase cycle and know exactly what you want. The mistake most marketers make is thinking everyone who arrives is ready to buy, so they jump straight into an aggressive pitch. My guess is eighty percent of people who come to your website are not ready to buy, so you have to build a careful, quiet relationship and hope you are around when they do have the need.

There is also the concept of triggers: it is hard to sell antivirus until someone has had a virus or read a scary article, or backup software until someone has suffered a loss. Think about the triggers that cause your buyers to realize they need your product, and, as a marketer, ask whether you can cause that trigger to happen, which is very powerful.

Top of the funnel: from a customer with no idea they have a need to one with a clear need, with a spectrum of gray in between. When you finish the top of the funnel, drive them to your website and get their email address, because without it you cannot stay in touch and build the relationship. Middle of the funnel: determine who is actually ready to buy and put attention on them with qualification, and put the rest into a lead-nurturing bucket, using software like HubSpot. The art of great lead nurturing is segmenting your customers, because generic emails have low open rates, but an email that says “I know you are a photography company, here is how we solve problems for photographers” does very well; similarly, if you track which product features someone used in a trial, you can email “I see you experienced this part, let me encourage you to look at this, here is a video.” Build a database of both the attributes of the person and everything they have done: were they on the pricing page (ready to buy), the technical pages (technical buyer), or the testimonials page (business buyer), and treat them differently.

Metrics for a funnel are simple: measure how many people are in each step and how many convert to the next. A common problem is that companies have not connected their activities together, so there is no link between the webinar they held and the next step; design a completely linked series of actions. For a simple funnel of visitors to trials to closed deals, track the number of visitors, trials, and closed deals (trend lines going up and to the right), the conversion rate from visitors to trials and trials to closed deals, and the overall conversion rate, because different lead sources (Facebook ads versus inbound) have very different conversion rates and paybacks. Every funnel, even Cisco’s, Microsoft’s, or Oracle’s, has blockage points; I spend a lot of fun time at companies finding them.

At a blockage point, two things stop people: friction and their concerns. As marketers we must create a motivation more powerful than the resistance and answer their concerns (sometimes just assuring them you will not spam them). JBoss (an open-source Java application server) had five million downloads and a business selling $27,000-a-month documentation and training, making a couple hundred thousand dollars a quarter, and a new idea to sell support contracts. I asked for the names of the five million who downloaded it; they had tried putting an email form in front of the download and it cut the download rate by a factor of ten, because nobody wants to give their email. After three months of arguing, they gave the documentation away free, which turned on a lead flow of 10,000 leads per month (too many for most salespeople to handle) and fueled the whole business. The free documentation was adequate motivation to overcome the concern about spam.

Website Grader (the predecessor to Marketing Grader) worked because it is a free application that gives a lot of value, so it spreads virally; it presents you as an expert and builds trust within seconds; and the score is a trigger, because if you are like most Americans and get a 55 (my site got a 95, but almost nobody does; typical is 50 to 55), you want to improve it, so you ask what step to take next. This concept is usable in almost any business: grade what you are selling, tell customers they are not up to best practices and why, and inspire them to act, especially if their boss might find out. Website Grader also shows the power of using engineering for marketing: developers can build far more valuable attractions than marketers, who are stuck with white papers and videos, so bring your engineering team in on this problem.

Go down to micro-steps. At Fetchnotes, a TechStars company with a simple to-do-list app, the founder wanted users to tag their to-do items, but people are not used to tagging and see no benefit, so the friction is that they do not know how or why. My suggestion: use in-app messaging that detects a to-do entered without a tag and pops up why they should tag and the benefits; after a couple of entries, show them the wow moment of clicking a filtered list (“get me my work phone calls” combining a work tag and a phone-calls tag), the moment they get excited and see the benefit. So micro-analyze each step of your trial: the friction, the motivation to move to the next stage, and especially the moment of realization that the trial was successful, so the customer will continue to use it and recommend it. Map your process from start to finish, take that one instant-buy website, figure out the mini-steps, connect each to the next, and write down the friction and concerns. The moment people write them down, magic happens, because they immediately see how to solve or simplify each one, whereas carrying them in their head they think they have done their best. My blog, forEntrepreneurs.com, has a lot more on this.

Outbound and the Hybrid Model (Mark Lorey)

There is no single right answer between inbound and outbound. One portfolio company started with a very inbound mentality and sold all the way up to $100,000-plus deals, even breaking the million-dollar mark, but at some point many argue you need a mixed model with outbound. Mark Lorey produced a document, on the site under go-to-market, outbound marketing, on the trade-offs and where outbound is applicable, so you can get the best of both worlds in a hybrid model. As Brian Halligan put it: if you are Salesforce.com or IBM with a big brand and you cold-call, people might pick up the phone, but if you are an unknown startup out of HBS selling performance-management software, that is an extra-special waste of time. Outbound starts to work as you get bigger; in the early days you hit the startup inertia. So there is a debate worth having, but the practical advice is that a hybrid “and” model is perfectly appropriate, and the best companies find exactly the right point to engage outbound from inbound.

The Play Card: A One-Page Go-to-Market (Mark Lorey, Spotfire)

We devised a tool called “plays,” using a sports analogy: it takes many players to get down the field. A play is how we go after a specific market segment, focused like a laser, executing cross-functionally with marketing, pre-sales, and even product changes. When I left Spotfire we were well north of $100 million, playing in several vertical and horizontal markets, so we needed the ability to go to market in each focused area, all flowing from the initial positioning statement.

Summarize your go-to-market on one slide. The play card for reaching risk professionals in financial services with our analytics product: on the left, who we are trying to reach, what keeps them up at night, the core differentiating message, and the competitive advantage over the next-best competitor. On the right, the tactical details to pull it off cross-functionally (creating demand, equipping the field and channels, enabling partners, product enhancements), so it is not a marketing-led initiative. Even a small company should get everybody on the same page, and you will be surprised what creative ideas come from pre-sales or engineering, who may know of a clutch you can address by building something into the plan.

On repeatability: we were going after different actors in different markets at different places in their buying cycle. The clinical-trials play delivered benefits immediately (we could win clinical-trial deals at will if I could get to the opportunity), so we measured everything. An awareness-generating play took months to show demonstrable results, so we measured top-of-pipe, watched things fall through the waterfall, looked for where people got stuck, and injected another customer tool or piece of website information to help them move down. Be patient with yourself: three months is not a lot of time to road-test your go-to-market unless you have very short sales cycles, and a stitch in time saves nine.