The most powerful change agent for early-stage startups: your assumptions
Expedite and de-risk the build, measure, learn cycle
Why do we repeatedly build products users don’t need or want? After partnering with 55+ teams in Google’s incubator Area 120, I’m confident there’s a better way to de-risk product development, one that promises more velocity, less resources and better outcomes. I’ve partnered with Barb, a design leader with early-stage startup experience, to write a series on this topic and develop processes that are repeatable.
Maya
Like the lean movement, we think one of the keys to better products is identifying and interrogating assumptions, but rather than validate through building, the maximal moment is pre-code. In this article we advocate to focus on assumptions and desirability if you want to define and build superior products, a process we call Assumption Led Definition (ALD). Beyond reducing risk, ALD is also a tool that facilitates understanding, drives alignment, helps teams prioritize, introduces a human-centered approach to early stage startups, and is an asset in determining the Minimum Viable Product (MVP).
The Assumption Queens
(Maya and Barb)
THE INGREDIENTS OF BELIEVING
Zero to one requires conviction. From cumulatively working with 94+ startups we’ve found believing in a product concept is inspired by a cocktail of knowledge, assumption, and presumption. When optimizing for velocity it’s rare to untangle and tune these individual ingredients. More often than not, knowledge, assumption, and presumption become an unspoken part of the product narrative that with each sharing becomes more “real.”
Accepting the assumption part of this trio as static is perhaps one of the biggest pitfalls. Powerful anchors, assumptions are the building blocks that inform a founder and their team’s perception of viability, feasibility and desirability.
Rather than take them for granted, our UX team has learned to lean heavily into assumptions to quickly de-risk startup concepts, evolve design mocks, and inform the product direction. Such investments aren’t one-dimensionally UX work, they are critical first steps to close the Product Market Fit (PMF) gap to assess desirability.
ASSUMPTIONS = THE LAUNCHING PAD
We’re not the first to broadcast the importance of assumptions. A key tenet of The Lean Startup is to validate assumptions which Eric Ries defines as “unproven beliefs you have about why your plan will work.” While there are plenty of great insights in Ries’ philosophy, his directive to implement the Engineering led sequence “Build, Measure, Learn” is costly. It doesn’t optimize for the true power of an assumption (and UX), which is to lead the iterative process and incrementally move towards PMF. Contrary to Ries, we suggest a sustained investigation of assumptions as the bedrock of our practice, which we refer to as ‘Assumption Led Definition’ or ALD.
Ignored, an assumption becomes a relic of the pitch or story and can drive confirmation bias, thereby limiting discovery, possibility, and value creation.
DON’T LOSE YOUR CONVICTION, UNPACK IT
From pitch to funding, beliefs are cemented and can be conflated with facts. Ask a founder “what are your assumptions”? They may have had a few in the early days, but once founders jump into build mode, assumptions tend to be expressed as knowledge.
Before we dive in further, let’s define the ingredients of belief:
Assumption: An assumption is an effort to connect and increase certainty. It’s a supposition made without evidence often informed by intuition, past experience, mental models, or norms. It requires research to understand if it’s valid or not.
Presumption: A supposition that can not be proved, in product development these are best abandoned.
Hypothesis: A statement that predicts a relationship between two variables which can be tested through collecting data in an experiment or live product experience.
Knowledge: Awareness or understanding of a subject or situation, gained through experience, education, or research.
Another trap founders fall into is repositioning assumptions as hypotheses. The appeal of a hypothesis is that it masks uncertainty and in theory can be tested. The danger of a premature hypothesis is that it commits the team to code and a solution based on unevaluated assumptions, meaning that potentially multiple phases of problem definition and discovery have been skipped. When testing a hypothesis the stakes are higher because you’re dependent on product interaction or usage (from a user you often know little about), and rarely can uncover the “why” of desirability.
The goal in de-risking an assumption is to transform it into knowledge. We’re assumption queens, here to reinstate their reputation. We hope you’ll soon see their innate power through the six steps of Assumption Led Definition:
Unpack and align. Start with a team-wide generative exercise. Map out knowns and unknowns. Unknowns often fall into three buckets: assumptions, presumptions, and hypotheses. A bit of structure can go a long way, categories we regularly prompt include: problem, market, need, value, experience, technology, business. Make sure to double-click into the assumptions, presumptions, and hypotheses. Often teams don’t have unified definitions, commonly referring to assumptions as hypotheses. As a result, ensure hypotheses are unpacked into the assumptions they’re based upon.
Bucket the assumptions. Once you’ve labeled everything in the map, pull out the assumptions from the presumptions, hypotheses, and knowledge. Organize the assumptions into the categories of desirability, viability, and feasibility. Pressure test the assumptions generated against business goals, OKRs, and functional roadmaps to make sure you’ve been as comprehensive as possible.
Prioritize by criticality. Now consider the criticality of each assumption. We typically ask how severe is this threat to our success? It’s not always easy to answer this question when success is 0-1; expect some healthy debate.
Focus on desirability. PMF is all about desirability. Assumptions that threaten desirability should be your first priority (P0s) because if they aren’t evaluated early, they could derail the entire business. It’s not worth building a product or figuring out its impact on the business, if ultimately no one wants it. The good news about desirability focused assumptions is that they can be assessed without building anything.
Evaluate market assumptions. You can evaluate desirability focused assumptions through a myriad of methods. We’ve used diary studies, surveys, and in-depth interviews. What’s imperative is how you pursue the evaluation. For example if you assume the product’s value proposition will resonate with a specific user, you first need to evaluate the pain or challenge as a baseline before you assess the potential benefit of the value propositions. It doesn’t need to be a multi-week investment, we’ve run research in five working days.
Refine the product: With your fresh learnings, go back to the drawing board. Does your target user have an explicit or latent need? Is your value proposition significant for the target user (i.e. if there a real pain or challenge you address)? Do you need to iterate on your storyboard or the concept? Does your MVP prioritize a compelling promise?
Repeat: Assumption Led Definition is a cyclical exercise. We recommend keeping a live tracker of assumptions to maintain movement. Adding a projection of potential learnings such as If we learn …. Then… helps keep an indecisive team accountable post-research.
STOP FAILING, ENGAGE YOUR ASSUMPTIONS
If you walk away with one takeaway…
We hope you befriend your assumptions. Don’t fear being wrong, instead realize the power of your assumption. ALD is a launching pad that could save you from wasting valuable resources.
90% of startups shouldn’t fail. Let’s stop shooting darts in the dark and throwing stuff at walls to see what sticks. Instead, adapt the go-to model of build-measure-learn.
The assumptions supporting your conviction are a good place to start. Create space for your product team to unearth and distinguish what is known vs. unknown. Start with learning. Pause before you invest in building and consider a more efficient approach to product development. Hire UXers who specialize in early stage startups (we know a few) to demystify desirability. Designers and Researchers are the partners you need to take a faster iterative approach (learn-prototype-learn-build-measure), with customers, with consumers, with prospects. They are the key to velocity. They will close the PMF gap.
 | A guest post by
|