Friend of Fuzy, Adam Dahlgren is the COO and Head of Product at Allstacks. Like Fuzy, Allstacks seeks to shine a light on hidden patterns––but Allstacks focuses that spotlight on the engineering team to better understand project health, assess risks, and create more accurate forecasts with AI-powered data analysis.
We sat down with Adam to get his perspective on how and why synthesizing multiple data streams to create rich and informative context helps technology leaders grow better teams and build better products. In our discussion, Adam emphasized the essence of Allstacks' and Fuzy’s overlapping ethos: outcomes data helps bring Product, Engineering, and the Business into alignment and catalyzes accurate, efficient decision-making.
But that’s all easier said than done. Here’s Adam’s take on managing products, the people who build them, and the data used to inform the work.
How do you think about measuring meaningful outcomes that can help drive your team, product, and business forward?
When it comes to outcomes for a PM, it’s important to partner with engineering to get good at getting small––producing a credible chunk of value that will move the needle for the people that you're building for. Before you can drive any real KPIs, any other outcomes, you need to be able to reliably deliver small, repeatable value increments. And teams should prioritize these increments (or features) based on the expected product outcomes. After that comes gathering feedback from the market and using that data to adjust accordingly. Without good data, and a way to unify and interpret it, you’re left eyeballing the market and using gut instincts to build this prioritized list. That’s why you see a ton of solutions aimed at making all of this more digestible and meaningful but most of them just scrape the surface.
Good decisions rely on context, and with current tools, data lacks that context or takes an insane amount of time to put together in a market where speed of innovation and experimentation are key. This is why tools like Fuzy and Allstacks exist. Some of the promise that Fuzy can fulfill is being able to more quickly and clearly isolate what’s really different from all the noisy interactions and day-to-day feed of activity that product managers can drown in. It’s a huge shortcut to have a feedback loop with insight on what really did happen so the PM can interrogate the data more deeply to find out what it means and what to do next.
As you prioritize and build these small value increments, how do you know they contain the user value you think they do? How do you know you’re choosing to build the right thing?
First, I will highlight that there are lots of different classifications of value–– there's user value, there’s customer account (commercial) value, there's internal tool value, there's learning value, etc. To me, those are all really still in service of end-user value. You should always start by defining what capability you’re building, who it's for, why we're doing it, and what we generally expect to happen. Nobody is perfect at this––and this is where something like Fuzy comes in to help you get better at it. To build alignment with engineering and get these features shipped, it's helpful if the team understands and trusts the impact it will have on the business, and in turn, their impact on the business.
I think a hallmark of really good product teams is not necessarily that everything always happens the way they thought it was going to. Sometimes being surprised and having to adapt is kind of the point. The market and your users will always surprise you. Your ability to see signals in the data and act on the feedback loops really determines your level of success.
What are the most significant challenges to harvesting outcomes data and how do you see the solution space addressing this problem evolving?
There was an era where the core data sources and activities took place in siloed tools that took many years to build and optimize individually––salespeople managing deals in a CRM, engineers writing code in a source code management tool. All those tactical things took quite a long time to be used across the board and create critical mass. Along the way, the industry evolved application performance management and then user behavior management. It just took a while to get to a point where all the data sources were available and ubiquitous enough in the market. Now you also have new better ways of intake, transiting, storing, and making sense of that data – plus the models that you can build on top of it. And so this evolution is very natural, but a lot of things had to come to pass for this to become possible.
In the case of Allstacks, we like to think about it in terms of having very low overhead and moving really fast to show answers and insights we can build on––a very short time to value. I think you'll see lots of different innovations as more insight comes from the merging of product usage and sales and marketing data. I think there's just any number of different flavors that come from that.
As a product leader, what do you consider the most important engineering outcomes to track? What do you wish you could track that so far has not been possible?
One thing I would like from my existing user tracking and monitoring point solutions that I think will become available with the promise of a product like Fuzy would be internal SLAs thresholds monitored by cross-functional teams to maintain a base level of health across accounts and ideal customer personas. Tracking adherence to these thresholds, combined with upstream marketing and sales data, would be super valuable. It would provide a more nuanced way to get the next derivative answer to user behavior tracking management. I think a lot of brands’ users are surprising them––or would surprise them if they were looking at it deeply enough. But the existing crop of user behavior tracking requires a lot of tuning and overhead. You have to pull out the insights; they don’t get pushed up to you.
There are a lot of ways to get insight across your software organization that can help you get better at delivering for your customers that never used to be possible––but not all of them feel modern, which is a term we use a lot at Allstacks. The solution has to be fast, intuitive, and actionable. I think people can tell the difference when they see it.
Expert perspectives from product leaders and data scientists play an important role in shaping the future of Fuzy and moving Product Science as a Service forward. We’ll be sharing many of the insights we gather from a variety of technology professionals on the Fuzy blog. Stay tuned for our expert interview series to deepen your knowledge in product science and borrow lessons learned from veteran industry greats and entrepreneurial up-and-comers.