Defocused crowd of people walking along a sunny city street

Lately I catch myself reading about AI, writing about AI, and even when I open an article about product development, most of it turns out to be about AI too. Which tools to pick. How to use them. How to build collaboration around them, analytics around them, validation around them, how to rebuild the whole SDLC around them. When I looked at this year's product management trend lists, that's essentially the whole menu. None of this is wrong - I contribute to it myself. But I started to miss the topics we used to discuss before AI tools took over the conversation.

The pandemic pushed many companies to expand their online presence, remote work, and a stack of online tools almost overnight. Then, a couple of years later, ChatGPT arrived in November 2022, and the models behind it had been in development for a few years before that. Forced online work first, then AI tools - and we have been living in that world ever since. That is why I find 2016 interesting: product management back then focused on different things. There were tool discussions too, but they had not yet been touched by what happened from 2020 onward.

So I dug up a set of 2016 articles - trend pieces, three surveys, and a podcast - and picked the themes that kept recurring in the ones I found. Treat it as a walk down memory lane. I don't think the 2016 topics were better than today's, and every organization faces different challenges, so I leave the judgment to you - maybe one of these older ideas is exactly what yours needs. The one that stuck with me most is at the end.

The profession was growing up

Mind the Product's 2016 survey drew over 1,200 respondents, roughly double its sample from three years earlier. The share of product managers reporting directly to the CEO or board rose from 49% to 56%, average salaries went up across levels, 49% received equity, and female participation climbed 7 points to 35%. Other writing from the same year noted that formal PM training options were multiplying - dedicated courses, university programs, meetups. Taken together, these sources make 2016 look like a point when product management was becoming more formalized and visible - not a starting point, but one stage in a much older profession's consolidation.

PMs wanted to get out of the weeds

Pragmatic Marketing's 2016 survey of more than 2,500 product people captured a tension that still feels familiar. Respondents estimated that 72% of their time went to tactical work and 28% to strategy, even though 62% said their strategic share had increased since the previous year. Their desired future was almost the inverse: 65% of their time focused on strategic activities. They also reported working 48 hours a week on average. The report's future-facing comments predicted more specialized roles and cross-functional, semi-autonomous teams. In other words, the profession was gaining influence, but the generalist PM was already carrying too much of it.

Roadmaps were escaping PowerPoint

ProductPlan's 2015 roadmap survey found that most product managers still managed roadmaps in PowerPoint and Excel - and were frustrated by it. The solution being discussed was the "living roadmap": a fluid document that gets revisited, reprioritized, and updated as customers react, instead of a plan set in stone. Purpose-built roadmapping tools were the emerging category - worth noting the survey came from one of those vendors, so they had skin in the game. Still, the tools largely won: in the teams I have worked with since, a roadmap living in a slide deck reads as a relic.

Agile was still a transition, not the furniture

A Mind the Product article from February described one enterprise software company's five-year move from Waterfall to Agile. The old model involved long specifications, handoffs, and releases every 12 to 18 months. The new one shipped monthly and adapted faster, but it also pulled the PM into an endless cycle of backlog prioritization, user stories, and engineering questions. The company first split the job between a market-facing Product Manager and a team-facing Product Owner, then reunited ownership of priorities under the PM and moved go-to-market work to Product Marketing. What reads today like an org-design debate was then a firsthand report from an unfinished transformation.

Data went from scarce to overwhelming

The 2016 complaint was not a lack of data but, as one article put it, a "firehose of data." Product managers sat at the intersection of streams from support, sales, finance, and IT, and the skill being discussed was turning that flood into decisions instead of drowning in it. A decade later, the firehose has not gotten smaller. We mostly just added new pipes.

But customer evidence was still missing

Having more dashboards did not mean teams knew what to build. In a small 2016 survey of 47 Mind the Product readers, 49% named proper market research and validation as their biggest challenge; among enterprise software PMs it was 62%. Several said they had no time for customer feedback because specifications and internal demands consumed it. The sample was small and heavily US and B2B, so it is directional rather than definitive. Even so, it exposes a useful distinction: behavioral data can tell you what happened inside the product, while customer research helps explain the problem worth solving.

The MVP was being rescued from its own name

One of the pieces I found reframed the Minimum Viable Product as an "experiment vehicle": build only enough to test a hypothesis while keeping the experience coherent. I would add a modern guardrail: apply the same standards for consent, privacy, and harm as in the finished product. Eric Ries's advice was to cut the planned feature set in half, then in half again. The favorite examples were the first Facebook - profile, search, friend requests, messages, built in a few weeks - and the first version of eBay, written over a weekend. The examples have aged, but I think the discipline behind them did not.

Products were becoming parts of systems

Another 2016 trend piece argued that B2B products should stop behaving like islands. As companies moved more work into SaaS tools, APIs and integrations became part of the product value itself, not just technical plumbing. The same article advocated automating customer workflows and using proprietary product data to create new value. This was a quieter shift than mobile or AI, but it changed the question from "Is our product good?" to "Does it fit into the system where the customer actually works?"

Growth became a job title

In early 2016 Harvard Business Review published "Why Every Company Needs a Growth Manager", describing a then-new role at the intersection of product and marketing: owning acquisition, activation, retention, and upsell, built on data infrastructure and constant experimentation. The evidence was concrete - Facebook's growth team had found that a key driver of whether new users stuck around was connecting with at least 10 friends in the first two weeks, and Pinterest lifted activation by more than 20% by redesigning onboarding. The role is common enough today that it is easy to forget someone had to argue for it.

The advice that aged best

My favorite find is from January 2016. Paul Adams, then VP of Product at Intercom, was asked what product teams should focus on in the coming year. His answer: ignore the trend lists. Without really good analysis and insight behind them, articles like "Top ten new technologies in 2016" are, in his words, a waste of space. Instead he suggested a "666" horizon: a six-year vision of how the world changes because of what you build, a six-month plan toward it, and six weeks of concrete work. He also liked a framing he borrowed from Ken Norton's Mind the Product talk - aim for 10x improvements, not safe 10% ones. I am aware of the irony of quoting this at the end of what is, structurally, a list of 2016 trends. His caveat was that such lists can be worthwhile when backed by strong analysis and insight - whether this one clears that bar is your call.

And here is the detail that made me smile. When Adams wanted to illustrate what a 10x improvement could look like, his example was to stop designing a settings toggle and "start thinking about artificial intelligence and machine learning" - a system that "just works by magic." In 2016 that was the far end of ambition, one PM's picture of a distant 10x future. In 2026 it is the default content of nearly every trend list I open - which, I suspect, is exactly why he would tell us to stop reading them and get back to our six-year vision.

Going through these articles reminded me how much of the job has nothing to do with any particular tool - roadmaps people actually believe in, data someone understands, experiments that actually test something, and a vision longer than the current hype cycle. Have you gone back to what your field was discussing a decade ago? I'm curious what you found.

Thanks for reading!

Sources

← Back to blog