Thanks!

Thank you for subscribing to our content. You will now receive the latest news in your inbox.

September 21, 2023

How Fast to Hire

“I think it’s working, but I don’t know how fast to go on hiring. How much burn is acceptable?”

Startup hiring is unintuitive because it is so uncomfortably nonlinear. Team often bloat ahead of strong product-market fit, but are usually too slow to scale right after.[1]

The Mirage of the Last Feature

How does this play out in engineering? You ship the MVP with a skeletal team, but sales are sluggish. Customers, request additional features, whether genuinely interested or merely polite. The team misinterprets this false signal as a sign that just “one more feature” will unlock demand. But this is often a red herring; the customer's indifference persists despite your Sisyphean efforts. If the core thesis is right enough, the customer will clamor for the product despite its gaps.

A larger team not only increases burn but also introduces inertia and resistance to strategic pivots. Founders feel pressure to make “progress” that reassures larger teams. They struggle to be intellectually honest, instinctively protecting the emotional state of passengers. Compound this with the overhead of management and communication as you grow – and the result? A high-cost, high-pressure environment that resists change.

Stay uncomfortably small until actual customers (not Silicon Valley spectators and “design partners”) are pulling the product from your hands.

Large Customers, Longer Delusions

This error of overhiring is amplified for startups serving large enterprises.  Long sales cycles, idiosyncratic wins, and great sales people can create illusions of partial success, obscuring the absence of genuine product-market fit. Customers will buy things for years that aren’t actually used if they are well-marketed.[2] Large customers will also put you in proof-of-concepts or the unending purgatory of “labs.”

For those serving the enterprise from the start, real engagement from customers, rapid timelines, conversion of POCs to paid contracts, and usage growth are all true-positive signals to look for. Consumer and prosumer founders are more naturally adept at being cockroaches while they wander in the dark, because they’re expecting high strength of signal (adoption, usage, virality, inbound interest). [3]

Of course, some products just can’t be shipped with 10,000 person-hours of engineering work. The minimum useful scope of many valuable products (Figma, Rippling, Arista, Nvidia) is just much larger than that. But this is the exception, not the rule. Most of the time, product-market fit is reached by iteration. Those who get to play the most hands have the highest chances of winning.

Venture Capital Has Nothing to Do With It

None of this is intuitive. Even worse, venture financing is occasionally the culprit for overhiring and failure to manage out duds. Venture funding is not a barometer of how many people will make your company successful, it is only an indicator of market demand for shares in your company. Your customers and your strategy should determine your fundraising, not the other way around. Whenever teams want to raise a very large early round, they say, “we’ll continue to spend like it’s not there.” However, the last few years are littered with examples of companies that raised an enormous amount of financing, grew huge teams ahead of product-market fit, and eventually collapsed under their own weight. Constraint breeds creativity and efficiency.

Phase Shift

Once you’ve built something customers want, a phase shift is required. You may be stable at a 6-person engineering team in product exploration mode, just trying to stay alive for several years. When you finally feel market pull, it feels reckless and alien to suddenly try to add marketing, sales, support and “business” people faster than your organization can digest. But to fail to do that is to limit your own growth.

Why is it so difficult to scale early go-to-market optimally? Engineering teams have never hired GTM before, so they often have false starts. What makes for a good engineer does not make for a good sales rep. Founders tend to gravitate to respected tech brands and domain experience above performance. If they’re selling a technical product, they can’t imagine that a non-technical person can sell it.[4] They fear the cultural change that comes with shifting from an R&D-only shop to a business, and sometimes it ends in organ rejection.

Hackers also love systems that scale. The manual, repetitive, demoralizing work of initial sales outreach is sort of culturally abhorrent to them, and they’re looking for efficiency. They dislike the idea of having as many salespeople as engineers. They often want to innovate on sales, and experiment with the idea of quota-less organizations. This is a mistake – incentives work. Get people who are great at GTM around you to teach you to hire.

Half of A Growing Sales Team is Unproductive

Uncomfortably, the majority of the sales reps, even in any working startup, are not productive. They are “ramping” or still learning the product, or not “making quota,” achieving their assigned sales goals. The former is expected, the latter is hard to disentangle from the fomer. This leads to two common problems – either the organization retains non-performing sales reps for too long, or the leaders in the organization want to fix “the system” before they scale it. Neither of these is ideal. One leads to a culture of nonperformance and high cost burden, the other leads to subpar growth. Great early stage sales leadership requires right-sized hiring with people who are able to succeed in ambiguous situations, figuring out the playbook as you go, and executing aggressive performance management with incomplete information.

Startups are Helicopters, Not Planes

A final reason for underhiring right after finding product-market fit is the challenge of keeping a growing, delayed-feedback, complex system in balance. You always either have too few qualified leads generated by too few sales development reps, or too few ramped account executives, or a broken sales process, or you can’t deliver change management to make your customers successful fast enough. There’s backpressure somewhere, but you have only lagging indicators to look at. Startups are more like helicopters than planes – they tend toward imbalance and require constant anticipatory adjustment, and if you wait too long to make change you will overcorrect.

The only solution is to get comfortable being uncomfortable. The first order of business is to stay alive long enough to make a product customers want. The second order of business is to, post-product market fit, sell as much of it as you can as quickly as you can. Early leads can be squandered, and the shift between the two is a crucible moment.

Notes

[1] There are two other very hiring-pace mistakes not addressed here (where we really focus on the phase-change around product-market fit) that are worth mentioning.

First is how hard it is to make the very first 1-2 hires beyond the founders. It’s scary to add folks that might change the culture, especially when the culture is two people who trust one another deeply, hacking in an apartment. It’s hard to imagine that anyone else could be as committed to the cause. But no company is built by two people.

The second obvious issue is bloat after the initial non-engineering functions are built out. Many startups scaled prematurely over the past 3-4 years, fueled by overcapitalization, false signal from enterprise customers buying nice-to-haves in a zero-interest rate environment, and COVID head-fakes of behavior change. While there's an entire science of scaling, the simplest solve here is to create a limiting function, something that genuinely reflects the cost structure/efficiency of your business (e.g. revenue or usage per employee), and to compare it to the companies you want to emulate. I’m optimistic that every function can scale significantly further for growing companies that harness the productivity benefits of AI.

[2]  Jive Software.

[3] More here about early testing of startup ideas for B2B.

[4] The successes of the sales forces at companies like Datadog, Databricks and Amazon falsify this belief.