“If we have data, let’s look at data. If all we have are opinions, let’s go with mine.”
— Jim Barksdale, former Netscape CEO
If you're fortunate enough to work in a vibrant startup, passionate debate is part of your everyday routine. Business people, engineers and designers all have valid perspectives, but their interests aren't always aligned.
Debates can get heated, so it's important to stay focused on the facts. That's where data points come in.
What are Data Points?
The term 'data point' comes from statistics, but it's taken on a new meaning in Tech. A data point is something you can point to. It's a piece of information that stands as objective evidence to prove your point, but don't be misled by the word data… a data point can be many things.
You don't need a lot of web traffic to start collecting data points. A data point can be something you heard at a party. It can be a sound bite you heard on the news. User feedback sessions and analytics dashboards (like Google Analytics or Mixpanel) are full of data points.
Data points have to be easy to cite in a product discussion. Businesspeople interact with customers every day and have experiences that are easy to reference. Engineers have objective data-driven analysis to call on. Often designers rely on feelings.
Sure, designers have color theory and UX principles, but in 2015 designers need to gather data points to hold their own in product discussions.
There are dozens of mediocre apps for each good one. Why? Because most companies are still driven by ROI (return on investment). It's easy to count dollars, but your product will suffer when dollars are your only data points.
Gathering Data Points
While data points can come from anywhere, there are a few reliable sources I can recommend.
Talking to customers is a great use of time for a designer. While structured user feedback sessions are essential, there's nothing quite like opening up a chat window and reaching out to users in real-time. It's immediate, fresh feedback about your product. If a user is frustrated, you feel it. If a user is excited, it's rewarding.
Analytics tell a story. Products like Mixpanel can tell you how a user got to your site, where they got stuck, and what platform they're on. If someone is complaining about your design, it's useful to know if they're accessing your site from a mobile phone.
Structured user feedback sessions
Online chat is immediate, but face-to-face time has a value all its own. It allows you to read the reactions on your user's face. You can watch them light up with a feature request and wrestle with a workflow.
I structure hour-long feedback sessions into three parts.
Part 1: Break The Ice
I always begin with an informal chat.
- How long have you been using the product?
- Does it save you time?
- What can we do better?
Broad questions like that can jumpstart the conversation. Let the user direct this portion of the session. Play therapist. Poke around to see what's on the user's mind.
Part 2: Observe
Spend the next 20 minutes watching the user use the product. Put on your lab coat and play scientist. It's tempting to direct the user with verbal cues when they get stuck but resist the temptation. Those are the exact moments you need to absorb and consider to redesign.
Part 3: New Features
The final part of the feedback session is spent on new features. This is a good time to present mockups and discuss what you're working on for upcoming releases. There's a natural opportunity to suggest sitting down and discussing a new feature when it goes live. Open-ended relationships with engaged users provide valuable insights.
Don’t Forget to Include Coworkers
Coworkers are a great source of data points too. The customer support team talks to customers all day long. They know what needs to be fixed and they don't mince words because they deal with customers when they are at their lowest.
Of course, if you're fortunate enough to work with a data scientist, they live and breathe real data.
Account managers have more established relationships with users, so they can provide an in-depth perspective on what users need. Good salespeople find you and force data points down your throat! Great salespeople push you to build features they've already sold to clients.
Gaining Authority with Data Points
It's always clear when someone has done their homework in product discussions. The worst thing to present is isolated anecdotal evidence. Anecdotal evidence is powerful, but a lone data point is not enough.
You need to convince your collaborators that what you're presenting is an established trend. If you only have one data point, you can float it as a suggestion. Sometimes it gets reinforced by someone that has come across something similar, other times it's just ignored.
Data Points in the Real World
At eShares, where I design and manage a product team, security is a big concern. We handle sensitive financial information, so a simple summary page can spark a passionate debate.
We call these debates "cage matches," because of the passion with which people argue their points. But don't let the name mislead you, these meetings are polite, fun and constructive.
When designing the page above, our in-house legal counsel voiced concerns about exposing the email addresses of board members. When asked for evidence, he recalled multiple customer complaints about privacy. He made his case with data points.
As I accepted what he was saying, I realized we had a layout problem. We needed to display email addresses for the People section.
How could we show emails for everyone but board members? It would look inconsistent, so I came up with a separate Board Members section and gave it a contrasting horizontal layout.
Now the design looks structured and consistent within each section. That's just one example of how data points have informed our product.
Who needs Data Points?
While these lessons are most valuable for product designers, freelancers and founders also need data points to be convincing in their respective roles. Working with a difficult client is made easier by presenting facts. Pushing back against a seed investor is acceptable if you can point to specifics.
Data points substantiate opinions and provide a way forward that everyone can agree on.