Summary: Incentive structures for product managers at major tech companies are tied to the wrong outcomes and result in bloatware. Once a product finds PMF - that product should cease to make major changes rather than feel forced to continue innovating and evolving unless absolutely necessary. A product manager should be equally rewarded for justifying making no major updates as they are for actually making major updates.
Let's say you want to ship a product. A typical bare-bones-no-extras team is structured as follows:
- Product Manager - works out what you're actually building/the core value propositions/oversees the whole thing.
- Designer - makes the wireframes, implements the branding etc.
- Developer - (can be split into back-end/front-end) - take the designs and PM directions and build it into an actual thing consumers can use.
Once the initial build is complete, the product is shipped and the product manager starts talking to all users, analyzes usage data and focuses on implementing feedback until the product finds product market fit (pmf). (Basically: PMF is when users in your specific market actually use what you have built).
Good product managers at this stage of a product are very valuable.
So what's the problem?
Once a product finds strong PMF, feature-creep should cease and the team should begin instead focusing on making small improvements to speed and usability. What almost always happens instead: the PM needs to continue to invent additional product features in order to justify their job title and keep their job.
In an ideal world, it would be enough to say "This product has PMF - let's stop making any major changes". Instead the organizational heads in so many corporations I have seen tie performance bonuses and KPIs directly to the amount of things that get shipped. Ship a new high impact feature? Great - you get a raise/good review. Imagine instead if a PM said: "the product is perfect and I am actively making decisions every day to keep it the same by analyzing user data and patterns and competitors" - that active decision to not build a new feature in should be rewarded just as much.
The impact of this is subtle at first but over time becomes huge: One update at a time, really great intuitive products evolve to become 'bloatware' until eventually they lose their userbase to a competitor that have sprung up with a 'lite' version of the same core value prop.
Some examples:
- iTunes. This product was used by absolutely everyone back in the 2000s. Everyone loved it. Apple added a dozen features every year that slowly took you away from the core value prop (literally just play music on your computer). It became super bloated and lost huge chunks of its userbase long before Spotify and streaming based competitors had really come onto the scene. By the time Spotify came out, instead of updating iTunes, Apple were forced to release an entirely new product (Apple Music).
- Jira. This issue tracking software from Atlassian was the market leader a few years ago when it was super straightforward to just build issues and track them in sprints. Over time, they added so many additional features that I've seen countless engineering teams switch to using lightweight competitors (Trello for instance - ironically also owned by Atlassian) or even just a simple spreadsheet to track issues.
- Apple Weather on Mac. The weather on Mac used to be a really delightful UX. You could click down to see individual days, scroll through it and interact to see what time the sunset was, see multiple cities in one neat view etc. A few years ago - in order to justify the need to continue building, Apple unveiled a largely useless widget feature for Mac and switched the weather experience to a non-interactive tile that shows almost none of the core value prop (telling you what the weather is quickly) and has no ability to expand. Even worse: If you click it, it literally takes you to weather.com, a site filled with ads and spam, just because you wanted more info on the weather.
- Flux. You might still use it - I don't know. It was the first popular 'I change your screen color at night so you see less blue light' desktop extension. Originally it was a great and simple experience, you simply set the color amount you wanted and it stayed set. Then they rolled out this complex update where the amount could not be changed to go above or below certain limits unless you literally changed the timezone on your clock. Who on earth thought that was a good idea. Myself like many others deleted it - they added extra steps to get to the core value prop because for whatever reason they felt pressure to continue to innovate something that had already achieved PMF.
- Adobe Products in general. (This example is simplified - Adobe lost to Figma and Canva for a variety of reasons, but feature bloat is definitely one of them). 10 years ago Adobe was king. Photoshop was super lightweight - you could fire it up fast and quickly make edits to a pamphlet or flyer or social post, render it and be done in seconds. If you wanted to edit a photo it was amazing and intuitive and worked super fast. Over time they added such an insane amount of features that by the time Canva and Figma sprung up - Adobe was perceived as bloatware in favor of Canva's lightweight approach to simple daily editing tasks. Hoping to dominate the 'professional' market, Adobe clung to their feature-creep, but the designer market shifted almost entirely across to Figma. Eventually, Adobe was forced to acquire Figma for 20 billion dollars. If Adobe had kept Photoshop super light and web-based and reacted to the market demands slowly rather than adding thousands of features per year and locking themselves into desktop from GPU requirements, this might not have ended up costing them such a huge amount of money.
All of these products would likely have faired better over time if they'd instead had a PM stand up to upper management and say: "Hey - now that we have PMF, my performance should be linked purely to KPIs that directly impact revenue and nothing else. We don't change this product unless we absolutely have to, we just make small minimal improvements across the board over time and play defense while watching what competitors build".
Here are examples of products that have remained market leader by not feature-creeping:
- Google Search. The home page of Google's search engine has not changed, ever. The same minimal layout brings you straight to the core value prop: search for something and get it. They have instead focused on constant speed and search-return improvements and have remained market leader in search almost entirely uncontested.
- Spotify. Spotify's UX is exactly what you expect: you open it and you see all music in existence organized in every way you could want. I can't remember the last time Spotify made any kind of major product change. They slide in occasional 'delightful' moments (pairing your account with a friend to see mutual tracks you both like was a cool one recently) but fundamentally the experience has largely not changed since they found PMF.
- WhatsApp. This app has 2.2 billion active users, and unlike other Facebook products it has remained largely unchanged in terms of the core value prop and the ability to get to that prop. The layout and UX has remained almost identical since it was first released and the userbase has only grown with time.
There are lots of possible solutions to this, but one solution is that we started recognizing that early stage and late stage companies need entirely different skillsets for a PM.
One possible solution: Break up Product Manager roles into "Pre PMF PM" and "post PMF PM". At early stage the product manager is focused solely on building something people will use. Once that is achieved, they take their skillset onto a new product entirely and palm their existing product onto a 'post-PMF PM' who excels at maintaining and iterating minor, impactful improvements over a longer time frame based on data and cautious competitor analysis.