The Little Things
(Re)making the case for experimentation & lean principles in building big businesses
It’s easy (and intuitive) to associate success with size.
Big visions drive big innovations, which, in turn, generate big teams, revenues, and businesses.
But, in my experience, it’s most often the little things — the little details, learnings, and iterations — that distinguish big success from failure. Even the best thought-out plans must be routinely tested & tweaked to retrofit strategy to reality. And at times, they should be turned on their heads to address unexpected shifts in culture, technology, and/or regulation.
I’m far from the first person to say this, but it sure feels like we’re losing sight of it.
Here’s what I mean:
1 — More & more early-stage capital is going into pre-product companies. This includes a bunch that are building on, or into, nascent platforms like the blockchain, quantum computing, VR, and even AI. These companies take longer to launch usable things, often limit customer visibility/feedback pre-launch, and bake in assumptions about the future with limited ability to adjust to unforeseen shifts. Many of these will surely make for good investments (have made some myself), but, as explained below, their effect on the startup zeitgeist at large concerns me.
2 — The lean startup method, once admittedly and unwisely elevated to “necessary and sufficient” status (it’s merely necessary, not singularly sufficient — more on this below), is going out of fashion faster than the mid-range jump shot or rap songs with 30+ words.
3 — Relatedly, demand validation (i.e. trying to establish early & cheaply that people actually want what you seek to build) is all but disappearing from early-stage fundraising. Many (probably correctly) are arguing that you’re now better off sharing nothing about validation/traction until you’ve cleared major revenue/user growth thresholds. And of course, if you’re not gonna share it, the likelihood of doing demand validation at all decreases as well — which, I believe, is to our joint detriment.
Experimentation (or “the lean startup method”) should not be the way you “find your idea”, or the underlying mission/purpose of your business. But, it is a necessary tool for iterating, executing and scaling an existing vision/passion/idea.
When thinking through these trends, I unexpectedly started to crystallize the moments in my time as an entrepreneur/investor where little details have ended up being necessary elements of big success. I thought I’d share them here — and see if I can’t help swing the pendulum back a bit closer to equilibrium :)
Example 1: Crisp white sheets (a $5 billion insight)
In 2013, I met a startup (then named Oravel) looking to build the first marketplace for affordable accommodations in India.
I’d spent enough time in India to know the intensity of the problem, and from a few Skype calls with the founder and one of his advisors — developed strong conviction in the team as well. But, the thing that stood out most to me was a small, but key, detail that could only have been uncovered by a deep commitment to experimentation & observation.
Even in the earliest days, Oravel had no problem acquiring hosts to supply its platform (there are plenty of vacant rooms/floors across India). They had also identified early adopters on the demand side (e.g. medical tourists on extended stays, young/new middle class vacationing with friends), but repeat purchasing was not as high as hoped. After a bunch of experimentation and learning, they established an incredibly important insight about repeat behavior — which eventually drove a key pivot in business model as well. Crisp white sheets.
See, not every host used white sheets (as most hotels obviously do) — but those that did got significantly better reviews than those didn’t. And that was a harbinger for a bigger insight. In a place as diverse as India, it was existentially important to create a reliable standard around this new class of accommodation — and not rely on shared culture/norms to create it. Fast forward five years and that same company (now known as OYO Rooms) focuses exclusively on owned/operated rooms, has become India’s largest hotel network and was last valued at $5 billon. Not insignificantly, the white sheets insight is still a key component of their value prop :)
Example 2: Evolving with your users (to reach millions more)
As I’ve blogged and talked about at length, the early days at Tala (then Inventure), were highly experimental. Importantly, all this experimentation was buttressed by a) shared belief in Shivani’s vision to expand financial access for billions of underbanked people worldwide, and b) deep commitment to this specific customer set. We wanted to solve this problem for that group of people — and the rest we would figure out over time.
Here’s, roughly, how that went.
1 — Shift in technology: When Shivani started working on the idea in 2011, there were virtually no smartphones among the underbanked in the developing world. By 2013 (when I started working with her full-time and going into the field in Sub-Saharan Africa/South Asia), we started seeing more and more of our target audience start using Androids. This happened a lot sooner than we expected.
2 — Shift in culture: Along with smartphone adoption, we started noticing two major trends in the way our target audiences used their devices. In places like Kenya & Tanzania, they used their phones to transact a TON. Second, we saw comparable apps use the Android API to access all kinds of data (including transactional) on users’ phone with simple/clear user permissions. This was a big shift because, up to that point, telcos & banks were true gatekeepers of this data in emerging markets and sought outrageous rents for it.
3 — Experimentation: Given these changing realities, we scrapped our plans and engaged in a heavy dose of experimentation around D2C data sharing that stayed true to our guiding vision/customer focus.
Pretty quickly, we found something that worked: an Android app that asked users for permission to access key data on their device and (based on an evolving algorithm) could qualify them for short term micro credit. Demand was through the roof — and, surprising to us, repayment (% of who paid back loans + interest) & repeat rates (% of repayers who took another loan) were extremely high from the beginning.
4. Little learnings/details: This early success notwithstanding, when we first launched this experiment, we had no idea if anyone would even take a first loan — let alone repay it and ask for another. As a result, we automated basically nothing and didn’t control for (in hindsight) obvious variables — like the speed with which you received a loan. Some folks got it within 30 minutes; some folks got it the next day. But this ended up being one of the smartest things we did — as it helped us establish a pretty clear causal relationship between the speed with which you received your first loan and the likelihood of repayment + repeat.
Fast forward four years, and Tala’s flagship lending app is a top-5 app in Kenya and the Philippines, serves millions of borrowers worldwide (with remarkably high repayment + repeat rates), and, yes, still emphasizes speed & convenience as key value props :)
Necessary but insufficient
Zooming out a bit, I think the key message from these experiences is that experimentation (or “the lean startup method”) is a neccessary-but-individually-insufficient element to building a great business. It should not be the way you “find your idea”, or the underlying mission/purpose of your business. It is an insufficient “why.”
But, it almost definitely should be a key component of the “how.” It is an incredibly useful & (I believe) necessary tool for iterating, executing and scaling an existing vision/passion/idea.
Amid all the white papers, grand promises about AI, and VC podcasts commanding attention these days, it’s become too easy to forget a once well-accepted reality. Without a habit of shipping quickly, establishing data-driven truths, and iterating accordingly — fully baked visions bucked by passionate founders can easily become enemies of the progress they seek to create.
- 100 page plans with 10 year financial projections (wild guesses at best)
- Endless diagramming on whiteboards
- Shouting matches over rivaling instincts on design palettes
- Losing sight of key blind spots (like new user experience & empty states)
- Obsessing over competitors (and endless intellectual pivoting as a result)
- External launch parties & addiction to press coverage
Man, that sh*t sucks. That sh*t happens all the time. And that sh*t will kill even the best of ideas.
But, a healthy dose of experimentation and lean principles — coupled with a firm & grounding in shared mission — will avoid it all. And, when done really right, that combo will the generate the few little details needed to create something big.
