While a technical background is a mandatory prerequisite for becoming a product manager, there are some technical skills worth having in your toolbox as a PM. The good news is you don’t need to go back to school to master these technical competencies either. The skills we’ll discuss in this article won’t put you in competition with your engineers or make you smarter than your system architects. But they WILL make you faster, more independent, and more knowledgeable about your product and your users.
Given the wide array of options for which technical skills a product manager could attempt to master, we’ve come up with a top-five that would burnish any product manager’s resume and are applicable to nearly every segment and industry. So without further adieu, here are five technical skills for product management you wish you already knew (and how to learn them.)
Data collection, extraction, and analysis
No product manager worth their salt would dare make an argument without having data to back it up. The problem, though, is it can be tough to get your hands on that data when you need it. That’s why the first technical skill we suggest all product managers master is data competency. Basically: you should know how to obtain, extract, and analyze the data you need to prove a hypothesis. Ideally, you’ll be able to do this without tapping a data scientist on your team on the shoulder for answers.
Learn how to wrangle a dataset confidently and you’ll be able to investigate hypothesis after hypothesis without getting on the nerves of your coworkers or requesting reports and extracts that may have dubious value.
Where to start: SQL
Regardless of how your data is stored, the best way to unlock it is by learning how to construct and run your own SQL queries. Not only will this allow you to run countless queries by yourself, but it might actually be a prerequisite for some product management jobs you’re interested in down the line.
SQL is not exactly cutting-edge
technology, so there is a vast library of independent learning opportunities out there. You can learn from the experts at W3Schools or take a course on Learning SQL or Analyzing Business Metrics from CodeAcademy.
There are also some resources specifically for product managers trying to understand what SQL is and how to get started on blogs from Product Management 101 and the Department of Product. Once you know enough to get started, you’ll need to request read-only SQL access from your engineering team.
While SQL will net you lots of raw data, making sense of it and visualizing it requires a different set of skills and tools. We’re living in the golden age of analytics, so there are plenty to choose from.
You’re probably already looking at Google Analytics by now to track web traffic and campaign conversions (and if not, you need to get on that pronto). But beyond that, Mixpanel, Amplitude, Looker, and Tableau are some of the popular products to turn raw data into gorgeous charts and graphs. Once you’ve selected a tool (or get a seat license for one your company already uses), it’s time to dive into their documentation and tutorials to figure out how it works.
From there you’ll be able to start churning out insight-laden visuals that can inform decisions related to feature prioritization, user behavior, marketing effectiveness, and beyond.
It might be 35 years old, but the classic spreadsheet program is still as relevant as ever. Sure, Google Sheets has taken some of the shine off this Microsoft Office standby, but there are a few things that still warrant product managers firing up Excel.
Where to start: Pivot tables and Macros
Pivot tables are the first value-add Excel has to offer for product managers. With them, you can quickly create a dynamic filterable chart or graph. Want to see how many orders this month were for Premium versus Basic? Then you want to show it in a graph? Pivot tables will do the trick in a heartbeat.
Macros are the other useful Excel feature a product manager might want to utilize. Using the VBA coding language, it can automate any task that you might perform in a spreadsheet, such as pulling data from various sources. Best of all, the “programming language” is all written in plain English, so it shouldn’t stump any neophytes.
Back in the dark ages, bringing a product to market required a series of educated guesses. There was no way to know if one option was better than another, so you’d simply bet on a horse and see if it won.
Today, we have the opportunity to continually introduce multiple options to the market in controlled experiments. We can scientifically prove whether the user prefers the blue button or the red button instead of just picking one and hoping for the best.
We run these experiments using the concept of A/B testing. Basically, some people see option A, other people see option B, we see which one got better results and then pick the winner.
A/B testing can be conducted on almost anything, from marketing campaigns to UX elements, to pricing models. Best of all, it provides data-based evidence to justify product decisions and foster stakeholder alignment.
Where to start: lightweight A/B testing
Now, the A/B concept itself isn’t particularly technical, but executing these tests does require a little more skillfulness. To ease into it, there are some tools that facilitate experimentation without any real coding.
Unbounce lets you test messaging, art, and calls to action by running multiple variants of landing pages, as well as pop-ups and banners on your core website. Tools like Optimizely enable experiments on your actual website or apps, randomly segmenting visitors so they’ll have different experiences. Based on the results, you can see which text, images, or product features are resonating better.
More serious experimentation
Depending on your use case, the A/B test you need may be more than cosmetic. To truly test out different variants of functionality or workflows, you’ll need to engage your engineering team.
When going this route, it’s essential to have completely thought through your experiment. As you’re asking developers to essentially code something that has a high probability of ending up in the rubbish bin, your plan must be bulletproof.
Start with your hypothesis and be sure it is fully socialized. Next, you’ll need to determine if two options (A & B) are sufficient, or if you need to get into a true multivariate scenario.
Then it’s time to determine what data you’ll need access to evaluate the results. Defining this during the design phase is key to making sure what you want is being captured, tagged, and categorized sufficiently.
You’ll also want to figure out how big of a sample size you’ll need to get statistically significant results. This will inform the overall size, scope, and length of the experiment.
Finally, you’ll want to establish what the outcomes of the test will drive. Do you simply go with the best performing option or do the results spur an additional round of testing?
Product managers don’t need to write, nor must they master UX design. However, it can be handy to have the skills to churn out a simple prototype for discussion or testing purposes.
There are a few reasons for this. First, sometimes you need some visual aids to convey your vision to stakeholders, and a rough prototype is a great way to make things tangible. You also might want to augment your requirements with a prototype to be sure the implementation team understands what you’re shooting for (even though any “design” coming out of product management should be purely informative and not prescriptive).
But, the most valuable use case is for collecting feedback. While surveys and interviews can glean all kinds of insights, sometimes you just want people to react to something. A prototype pushes things out of conceptual and provides a frame of reference for furthering the conversation. If you’ve whipped up a couple of options, you can see which one people prefer to better guide your plans.