Taming your assumptions mammoth I

Aditya Mishra
5 min readOct 20, 2020

A guide to process signals better and make stronger assumptions

image courtesy pixarpost.com

Assumptions Refresher

If you’re working in product you are heeding to Signals all day. Customer interviews, Usability tests, Competitor updates, stakeholder feedback, Customer Support tickets, you name it. You process these signals, generate ideas and then work on iterating , validating , shipping them. You’re driving towards outcomes, and these signals are simply the first step of many in your Product Development engine.

The signals you process leads you to create ideas and plan actions. These ideas and actions have desired outcomes. That your plan will result in desired outcomes are simply assumptions.

Two factors lead to the success of products

  1. Strength of Assumption Management: The process by which you create and validate assumptions
  2. Strength of Execution: The practice of you bringing in tech, design and operations together to build, ship and track your assumptions.

Great products are built when strong assumption engines meet strong execution prowess. Let’s distill this into a PM equation

Strength of Assumptions*Strength of Execution = Quality of Outcomes

Note: Assumption strength has to do with creation as well as validation. However a lot has been written about both hypothesis validation and execution. Very rarely we come across the assumption creation process itself and I’d like us to focus there. Thus the rest of this article dives deeper into our assumption engine itself.

Signals in the Product System

To understand how strong your assumption engine is, think of it as a system. A very crude PM system simply is a receiver of signals and a producer of assumptions.

In a perfect world: you listen to clean signals from your customers, make the perfect assumptions, have a great execution engine and thus ship successful product updates week after week.

This is typically true for few early stage startups. Co-founders savvy about the market and customers form a set of assumptions influenced by direct market inputs. A small group of passionate people building for a small set of users early on probably has the richest signal strength unaffected by extrinsic noise or distortion.

Naturally as teams grow and problems get deeper, product teams start getting signals from many sources:

Somewhere along the way you start listening to these other sources. You have reviews with your executive team and get feedback or new requests, your sales teams tell you things they are hearing from the market, your customer support team reaches out with issues that users have been facing, your data team has information you decipher insights from… (the list really is endless).

You start to pay attention and listen to different signals every day. Naturally, you perceive some signals more frequently and louder than the others. Some signals are heard the loudest, some get weaker. Each signal now has a “multiplier” associated with them which determines whether they are getting stronger reception by you or weaker.

Your original signal S0 (direct from the customer) gets diluted by the multipliers and signals from other sources. Thus the final signal that you end up processing is a combination of all the incoming signals.

Signal-Final= {M0*S0 + M1*S1 + M2*S2 …. }

You develop your own signal processing vector- a set of multipliers that end up determining the strength of signals from each source.

These multiplier coefficients form your own signal Vector = [M0, M1, M2…]

Understanding these coefficients for yourself and your team is the first step in identifying how closest your assumption is to your market.

Two important points to note about signal vectors

  1. Each PM has their own intrinsic signal vectors they’ve built over years of product practice and learning. Some PMs lean higher on direct customer signals, some rely too heavily on leadership direction, others rely on sales. Changes to intrinsic vectors are slow and need deliberate effort. This is a function of their past roles, product success or even natural inclination.
  2. Every product team has their own processes and culture that create an extrinsic signal vector. These are guided often by leadership, senior management or long standing individual contributors who have directly or indirectly shaped it.

Thus in practice, every PM uses a combination of their internal vector and their team’s extrinsic to create assumptions.

The in-practice signal vector comes from your intrinsic as well as your team’s extrinsic vector

This signal vector determines how signals reach you and get fed into your assumptions engine. Good product teams consistently strive hard to maintain loud customer signal multipliers , while still balancing the other multipliers. Maintaining a high customer signal multiplier is harder than it looks and only consistent effort to carve out time + resources can keep that number up. Sometimes you do come across teams that misinterpret other signals for direct customer voices. Then there are the most dangerous teams — the ones often great at disguising other signals as direct customer voices.

Being conscious about your signal vectors helps you make judicious decisions around product progress. Weak vectors always lead to riskier assumptions and need to be compensated with other means including relentless validation, thoughtful KPIs and experimentation.

Maintaining strong intrinsic and extrinsic vectors are not enough on their own to help you generate strong assumptions. The biggest reason being signals themselves are often distorted as they are transmitted and received. This is the “noise” component.

Part II Continued here



Aditya Mishra

I love drawing connections from different subjects in a hope to simplify the world of product management. https://www.linkedin.com/in/adityarsmishra/