"Everything Starts Out Looking Like a Toy" #58
On researching your hypothesis for feature design (finding what people need)
This week’s toy: robots that make your salad. While this sounds a bit like Rosie from the Jetsons, the ongoing focus on delivery and quick service food makes in-kitchen innovation a necessity rather than a toy. Investment there does not just lower labor costs, but also opens up new areas for menu innovation. Edition No. 58 of this newsletter is here - it’s September 5, 2021.
The Big Idea
When you’re thinking about a new market – or helping a potential customer enter a new market – a key concern is understanding what they want, and comparing it to what they need to achieve their goals. New features are tasty, and maybe not what the customer needs for the long term.
Creating the conditions for research success
“If I had asked people what they wanted, they would have said faster horses.”
– Henry Ford did not say this
It sounds really easy when you read the history of product success. The brilliant founder isolated the key elements of a customer need, hid in a secret laboratory, and emerged with an iconic product that people remember for generations.
The reality of feature and product discovery is a bit different and is much messier. You start with an idea, and then find that some customers have no overlap with your idea at all. Other customers might have that exact need and pronounce it in a different way. The key output here is to get a reasonable overview of the problems customers face and the way they describe those problems.
I loved this summary of product research by Jaryd Hermann, which emphasized:
creating hypotheses (what do you want to know and how are you trying to find out)
building a consistent way to inquire (create a list of questions you ask everyone)
being skeptical of an easy solution (what seems obvious … might be obvious)
moving forward when blocked (don’t analyze forever)
and delivering (shipping) the answers to those hypotheses (iterate, and do it again).
Jaryd’s insight is that product research works best when you don’t assume you have it right, but you do have a strong opinion that you need to prove or disprove.
Turning Research Success into OKRs
Once you have a strong opinion, validated by customer stories, you need to turn those needs into tangible things your team can do. Knowing which stories are big asks and which ones are simple items to tuck into a release is a key PM skill.
And the way we describe these tasks is changing. Instead of black and white tasks (you know exactly how to ship a feature from the getgo), there are more exploratory tasks that are entering the product process from the design world.
Compare the “use case” or “job to be done” approach with a design sprint. The former is a list of tasks when you really know your market. Perhaps you have a small improvement to make and the steps are clear. A design methodology, in contrast, might propose a few ideas and not know ahead of time which one will win.
Think of agile vs design as different styles rather than completely different philosophies. Each are systems trying to uncover the thing a customer needs. And they shine in different situations. You need to use elements of both to turn customer insights into features customers will use.
What’s the takeaway? Researching use cases that customers say they want is not always the same as researching the cases realize they need. When you start with an opinion on how to solve a problem, test whether users have that problem and whether they want to solve your problem in the way you’ve prescribed, you have a test that can be proved or thrown away. Iterate on these cases and you’ll get a great set of things to test with customers. But don’t analyze too much - you also need to ship to test whether the features will be used.
A Thread from This Week
Twitter is an amazing source of long-form writing, and it’s easy to miss the threads people are talking about.
This week’s thread: how much is climate change making hurricanes stronger?
Prof. Katharine Hayhoe @KHayhoe"Was it caused by climate change?" is the most common question when we hear about an extreme event. But when it comes to hurricanes, that's the wrong question. The right one is, "how much worse did climate change make it?" (thread)
Links for Reading and Sharing
These are links that caught my eye.
1/ Predicting the future - When you look at the history of graphs, you might not realize that some of the more useful methods of visualizing data are not that old. Pie charts and scatter plots have been around a scant few centuries (ok, it’s not brand new, but definitely not as old as some others). They are key in underpinning the kind of learning we are doing in machine learning and automated prediction today.
2/ Images of Roadside America - Public Domain Review highlights the photographs of John Margolies, chronicler of such weirdness as restaurants shaped like Giant Fish and others.
The bigger picture here is one of an America that increasingly does not have as much regional culture and weirdness. If you can find a big box retailer everywhere and a quick service restaurant that always makes coffee the same way whether you are in Sheboygan or Shanghai, you might have lost something important about being American.
3/ Becoming A Nation of Hermits - The Pew Research Center has found that attitudes among Americans have shifted with respect to where they live and what type of communities they prefer. Compared to before COVID19, Americans want to live father away from their peers, even if it’s less convenient.
On the Reading/Watching List
It’s time for Season 10 of one of those shows I can’t stop watching. Call The Midwife is back.
Data integration is one of the top responsibilities of business technology workers. The team at Workato found that 53% of full-time business technology professionals (534 surveyed) said data integration was their most important data task. Those same people don’t necessarily have control over building and improving that capability in their own team. (Only about 1/3 of those business operations peeps build their own automations).
I’m reading this survey to get more insights on what teams are doing to improve automation. (Conspiciously missing here is information how those teams document and share the work they do so they get the credit where credit is due for doing the work.)
What to do next
Hit reply if you’ve got links to share, data stories, or want to say hello.
I’m grateful you read this far. Thank you. If you found this useful, consider sharing with a friend.
The next big thing always starts out being dismissed as a “toy.” - Chris Dixon