Midwest startup tour: Little cities, no little plans

On Jan 23, I kickoff a five day tour of some of the Midwest’s most innovative cities and campuses. It seems fit to frame this with the oft quoted “Make no little plans” from our favorite son, planner and architect Daniel Burnham. He turned the railroaded cow town of Chicago into a model of urban innovation.

Silicon Valley is unequivocally this same model for startups and tech innovation globally. Its concentration of talent, risk taking mantra, capital and history make it one of a kind. Unfortunately, the Valley’s bending gravity leaves most investors blind to the forming critical mass of these same ingredients throughout the college towns and small cities in the Midwest – many of these places are becoming mini tech hubs. Why here?

Positive selection of people: 20% of the top 50 computer science programs, 26% of the top 50 med schools and nearly 30% of the top 50 b-schools are in the Midwest (US News). Startups are built on people, and like Silicon Valley, these programs draw the best of the best people from around the world. These are smart, driven people, willing to take risk.

Ideas are championed – nothing is impossible: The academic spirit aflame around these top academic institutions engenders both ideas and respect for ideas. This is a key ingredient as every startup begins with a founder’s idea and a few other people saying “wow, that’s a good one”.

Midwest values:  However, without Midwest practicality, ideas would die on the academic vine. The Midwest values of sensibility and action-over-words drive practical application and progress. They ask of any idea “So what? Show me”.

Abundant talent: The institutions of the Midwest yield a wide and deep talent pipe – a few of whom become founders – but many more who can join an already seeded startups. While the job market here is healthy, it is not nearly as tight or as expensive as Silicon Valley. Talent is more affordable in the Midwest and yet itself can afford a better quality of life. This yields a more loyal talent base with less turnover.

It is true that for rapid scaling a startup will need to expand beyond a single college town or maybe even the boundaries of the Midwest. In any given Midwest city, the talent pool may be high quality, but is also usually small in absolute size, presenting denominator challenges as a startup gets large. There are also fewer people here who have built large tech companies – the startup DNA is thinner, though deepening by the year – and capital remains relatively scarce. When these considerations are balanced well with the upsides, however, big wins happen.

Most know about Grubhub, Groupon, Braintree in Chicago, but think also of ExactTarget in Indy, Workiva in Ames, Duo in Ann Arbor, Toa in Cleveland, just to name a few examples. There are many more to come. We are already thrilled to be part of FourKites and G2Crowd in Chicago and Farmlogs in Ann Arbor, a truly a unique startup hub.

There will be more big successes, and we are on the road to find them. For founders founding and others looking to join a startup, below is my schedule so far. Email me if you want to meet:


Mon, Jan 23: @ Purdue in West Lafayette, IN

Tue, Jan 24: @ The Water Council in Milwaukee, WI

Wed, Jan 25: AM TBD, PM @ University of Chicago

Thu, Jan 26: AM @ Notre Dame in South Bend, IN; PM at U of I in Champaign, IL

Fri, Jan 27: AM @ U of I TEC Center in Champaign, IL

Next year already happened, were you ready?

Fred Wilson’s blog last week on end-of-year planning was a good reminder to get everything in order (strategy, plan, people) to hit the ground running in 2017. If you have anything of a hiring or sales cycle, however, most of 2017 is already baked.

End of year planning can work well for consumer businesses, where (once a product is developed) the turnaround time from spend to revenue and margin is very short. You can move the levers of marketing spend to drive more site traffic to conversions to transactions quickly to impact revenue – almost in real time. But even in these businesses, there persists a bottleneck on hiring people. The national average time to fill a job is about a month, though weighted heavily to hourly positions that are much faster to fill than professional roles. Startup hiring cycles are often longer. Then you need to add ramp time. Sales person ramp time, for example, is combination of training, experiential learning and the normal sales cycle baked together.

Here are rough guidelines on how long it takes new hires of different types to reach 100% contribution in a startup:

  • Executive: 6 months (many would say longer)
  • Developer: 6 months
  • Inside sales: 3-6 months (for the ~50% that make it)
  • Outside sales: 6-12 months (for the ~50% that make it)
  • Marketer: 2-4 months

So, in a consumer startup you might still have hiring cycle + ramp time = 1 month + 3 months = 4 months lag if you are relying on new marketing hires to scale the business. That means trajectory through Q1 to early Q2 is pretty baked in the worst case.

B2B businesses are much worse with lag time = hiring cycle + ramp time for sales people.

  • In SMB = 1 month + 3 to 6 months = 4 – 7 months
  • In Enterprise = 2 months + 6 to 12 months = 8 to 14 months

Based on this, if you are a B2B enterprise startup, your 2017 is pretty well baked already. Note that this math doesn’t include any capital raising needs which can add another 3-6 months if you need the capital to make the hires. Yikes! Given such a lag in cause and effect, end-of-year planning is not a very practical time for fast growing startups to set the plan for next year.

Instead, startups should maintain an 18 month forward plan at all times to avoid the hiring and ramping bottlenecks discussed above. This is equivalent to driving a dark road at night where your headlights illuminate a continuously rolling path ahead. This plan is a weekly referenced operating plan for startup management and something the Board should review and adjust formally 2-3 times a year. At the operating level, it changes all the time. With this plan continually in hand, end-of-year planning is more about how you will measure success and reward performance in the next calendar year; you already know “the 2017 plan” and are likely thinking ahead to 2018.


Bend the curve with Average Contract Value (ACV)

SaaS entrepreneurs know how hard it is to scale their companies from zero to their first few $M in revenue. I’ve posted before about misconceptions about scaling SaaS businesses – namely that adding more sales pods will lead to exponential growth. While pod additions are part of it, the single most powerful lever in driving exponential growth of a SaaS business is increasing Average Contract Value (ACV). I’ll assume for this article that ACV also equals annual contract value.

The other day our team was reviewing portfolio performance, and I nearly fell off my chair when I saw the power of ACV in the chart below (scale, years and a few data points adjusted to protect the innocent):

SaaS startup example revenue and customer growth


While after Q2 2013, monthly customer growth of this company stays relatively constant on average (meaning customer count grows linearly), revenue is a hockey stick. Why? Incredible gains in ACV of new customers.

The chart below distills this in more detail. In fact, below, monthly customer additions do increase but go up only a bit more than 2x from a relatively steady average of ~7 in 2013 to ~16 in 2014. However, new ACV grows from $5K in the early days to nearly $25K in the most recent months shown, a 5x increase. WOW!

New customers and ACV per monthpic 2.png

For those of you who like calculus, the 2x and 5x represent multipliers in a sort of second derivative of the ARR revenue level – the change in the change that bends the curve. In other words, over a period of a few years, the company is adding 2x * 5x = 10x per month more in ARR than it did early on. Yet it is doing this while only adding twice the number of customers per month than it used to. Assuming sales people remain as efficient from a # of customers per month perspective, the company is doing all this with only a 2x the size sales force. In fact, this company added more sales people than that because higher ACV also comes with somewhat longer sales cycles and higher levels of support. However, even if sales person deal efficiency is half of what it was, overall ARR efficiency of the sales force is still up 5x * 50% = 2.5x. I would invest in that all day and night.

So how did they do this? (you can do it too)

Mature your product (fast): A SaaS MVP usually lacks integrations, enterprise functionality and other controls and features – just enough for small to medium size businesses to find value in. To increase ACV, you need to give something to get something.

Who are the key adjacent software players in your ecosystem? More importantly, with which ones have your early customers expressed a desire for integration? Then there are the common MB and enterprise asks: single sign on, administrative controls, reporting, etc. Once you integrate a few key adjacent platforms and basic enterprise functionality, you suddenly have a means other than number of seats/users to drive price. Want Salesforce integration with SSO? Sure, that’s our “pro” version for 2x the price.

You don’t get what you don’t ask for: Early on, the startup game is about getting a few customers, any customers. You don’t really care what your early customers pay. Often they are friendlies and have agreed to iterate with you on product and feedback, so you’re fine with giving them a deal. Unfortunately, this can anchor you and your sales team to bargain pricing. While being conscious of competitor and comp pricing in your space, take off the gloves and ask for more. You need to prove that you can do this for yourself and your sales team. If your ACV is $10K when you hire your 4th, 5th and 6th sales rep, it’s likely to stay that way until you prove to them it can be higher. The only way to break the habit is to land some higher ACV sales yourself – as a company leader – and set a new precedent. You will both show your sales team that it can be done and get mad props.

Puff up your fur with killer marketing: Animals puff their fur in a fight to make themselves look bigger. Smart startups do this too. Our best SaaS companies are spending surprisingly large parts of their budgets on sponsoring key industry conferences. It’s how you look bigger than you are, help customers overcome startup aversion and grease the sales skid with name recognition.


Time to Heal, Learn, Unify

Like many of you, I was surprised by election results.

As we each individually decide what is next, what we do next, we need to see this in the context of 250 years of peaceful transitions across all imperfect leaders. We are so fortunate to be upset or ecstatic at the functioning of democracy rather than suffering war, coup or famine as in so many other places.

Now we – in our fortunate state – need to heal, learn and unify.

Heal: When I suffer loss, I find an immediate passing of time to be the strongest salve. Last night a low, then sleep, this morning fellowship in absorbing the discontinuity – whether you think it good or bad. By noon, back to business and the needs of our companies/customers, investors and team. Then sleep and more normalization tonight. The half-life of despair and joy in an otherwise good life (and we are all very lucky in the tech world) is short. By natural selection, our discount rates are too high to get stuck in any one experience in time. This is also known as hope, and hope is our greatest mechanism for survival and putting one foot in front of another.

Learn: We recently suffered a setback in our portfolio. That goes with the territory, but it got me thinking about the failure modes of startups. CB Insights has aggregated 166 post-mortem blogs posts and exposés from CEOs of failed startups. I read most of them. Yes, there are the obvious primary modes of failure: execution problems, PMF failure, no market, bad business model. But there is also an observable meta trend – high functioning people internalize blame amidst a search for truth. This is a powerful combination.

Today, if you don’t like the election result, what could you have done differently. Donate? Phone bank? Did you do those things? If not, do you have a right to complain?

More importantly, while some on either side may be “deplorables”: racists, misogynists… or crazy liberals, I am loath to cast stones. History will judge all of us as it inevitably does. In vast majority, millions of good people voted for both sides, including Trump. Outside of our ivory tower of high education, high class tech – where talent of a failed company is reabsorbed by the job market in mere weeks – millions of our fellow citizens feel left behind with limited education, diminished prospects and stagnant real wages.

What are we doing to help them? It’s our right and common mode not to do anything for others, but then do we have a right to complain about how others vote? “Of the people, for the people and by the people” means everyone, so this is on us too, on all of us.

Unify: The greatest signal of the strength of our democracy is an outpouring of calls for unity among leaders and citizens. As many continue to mourn or celebrate internally, we need an intentional period of unity to give a new government the chance to embrace all constituents. Partisan rancor early in a term, or before one starts, only sets precedent for an ineffective future and gives both sides excuse for irresponsible and non-inclusive governing.

If you are a democrat hug a republican; if a republican, hug a democrat. We all get the chance to fight again in four years. And remember, there was at least some good news for all of us last night. More states voted to legalize pot, and we get another 4 to 8 years of Alec Baldwin on SNL. Some levity, especially in combination.

Punch, type, click, swipe, gesture, speak, THINK

The user interface (UI) has markedly changed in the past few quarters. From “mobile first, native vs web and responsive” story lines, founder and investors’ dreams now echo the next revolution of UI, voice (or is it gesture?).

The chart below defines a recent history of internet/human interface device sales, starting with broad adoption of the keyboard + mouse via PCs in the 90s.


Source: PCs, Smartphones – BusinessInsider, Tablets – Digitimes, Echo – Meeker/KPCB, various

The transition from Static Epoch to Mobile Epoch coincides with an explosion of smartphone and tablet sales. This is not only a “mobile” story, however, it is a UI story as well. The mobile explosion was enabled by the development of small/cheap touch screens. Suddenly you could be connected and do anything anywhere with your fingers alone. Lugging a keyboard and a mouse around – or for that matter, typing on a Blackberry keyboard – were just inadequate for seamless mobile work. We are now entering a new age, moving from the Touch Age to the next. The question is whether it is voice or gesture or both.

Both voice and gesture have been around for a while. Most premium handset manufacturers have had voice (Siri, Google Now, etc) and gesture integrated in smartphones since 2013, and Kinect has been around since 2010. While these applications of voice and gesture had highly anticipated launches and early adoption, none proved to be mainstream. Few people I know talk with or gesture to their phones, and no one I know – save for a few overgrown gamers – own a Kinect. Tech cheerleaders would say the underlying technology had not yet been perfected enough for prime time, with burdensome training needs, reliability challenges and high battery use. However, my time as an engineer taught me that failed UI launches usually come from a lack of user insight rather than technology gaps.

It is simply weird to gesture or talk to my phone in a busy elevator or sitting among my colleagues at work. For kicks, I recently and unexpectedly “okay, googled” my Nexus 6P to setup a calendar invite, while sitting near my four quietly working colleagues. When I afterwards asked them on a scale of 1 to 10 how annoying it was to listen to me (10 being insanely annoying), they said 2, 6, 7, 4. That’s pretty annoying.

Context of use is everything. Mary Meeker’s 2016 report cites that 43% of voice search occurs in the car while 36% occurs at home. That doesn’t leave a big slice for work and social pursuits. Not so surprising; google glass wasn’t much of a hit around other human beings either, but the car and home do make sense as environments where social challenges are minimal.

I don’t believe either voice nor gesture will become as ubiquitously horizontal as touch, but they will both play big roles in parts of our lives. Here is how I see it:


As you can see above, I am more bullish on voice. Outside of VR (I broadly consider VR to be part of gesture), gesture is mostly a natural extension of touch with a few more degrees of freedom and planes of movement. For many use cases, touch is simply easier and more efficient than gesture. Voice, on the other hand, allows you to be in control when your hands are otherwise occupied or aren’t near a device, opening up a far broader range of new use cases.

Platform domination of gesture and voice a risk for entrepreneurs

The “so what” for entrepreneurs is that the tech giants are rapidly assembling their claims to be the voice and gesture platforms of record. The impact of this to startups depends on whether you are a pure voice or gesture technology or a voice or gesture enabled application.

Launching a pure gesture or voice technology is a tough road for startups because it means long sales cycles selling an embedded software to very large companies like Samsung, HTC, Google, Apple, Moto etc. Alternatively, it means a direct-to-consumer hardware device like Thalmic Labs’ Myo Armband. They seem to be doing well for now, but few startup hardware devices have happy endings.

Soon every B2B and B2C software company will be considering its voice and gesture enablement strategies much as companies endeavored upon mobile strategies en masse 5 to 7 years ago. This may evolve analogously to the iOS and Android app stores. Early in their lifecycles, app stores were a meaningful and sometimes differentiated distribution strategy for a software/app company, at least enough so to get viral adoption from an early base of consumers and hungry investors. No more. There is too much noise, and the data now show the vast majority of people spend their time on just three apps. The same will happen on Alexa and Oculus. Alexa is particularly problematic in that without any visual cues, I won’t remember to use that niche birthday reminder skill I downloaded last week. So, should you launch an Alexa skill or Oculus app? If they channel a UI that your customer demands, yes, but as a business model unto their own, no.

I feel you hearing what I’m thinking. Yup, gesture and voice are not the end game. It might be 10 years, it might be 25, but UI by thought is the final frontier. Companies like Muse and Neurable are in the earliest stages of commercializing this possibility. In the mean time, we have a few more years to keep our thoughts to ourselves.