Research Fundamentals Research Fundamentals Part 1: Research Before Testing: The Six-Step Process
Research Before Testing: The Six-Step Process
Research Before Testing: The Six-Step Process
⏱️ 2-min read
Bad testers jump straight to testing. Good testers research first.
Research tells you what’s actually broken.
Testing tells you if your fix works.
Skip research and you’re guessing. Do research and you’re solving real problems.
The Six Research Methods
Here’s the complete research process. You don’t need to use all six methods for every test, but knowing all six makes you dangerous.
Method 1: Heuristic Analysis
Look at your page through the lens of experience. What seems unclear? What creates friction? What distracts from the main goal? This is educated pattern recognition, not wild guessing.
You can audit a page in 30 to 60 minutes using heuristic analysis. We’ll cover the exact process in Part 2.
Method 2: Technical Analysis
Check for bugs. Does your page work on every browser? Every device? Does it load fast? Are forms functioning correctly?
Technical problems kill conversions silently. A broken form on mobile Safari costs you money every single day until you catch it.
Method 3: Web Analytics
Where do people enter your funnel? Where do they drop off? What’s the conversion rate at each step? Which traffic sources convert best and worst?
Analytics shows you the what and where. It reveals the size of problems so you can prioritize which ones to solve first.
Method 4: Mouse Tracking Analysis
Heat maps show where people click. Scroll maps show how far people read. Session replays show actual user behavior. Form analytics reveal which fields cause abandonment.
This method shows you how people actually interact with your page, not how you think they interact.
Method 5: Qualitative Surveys
Ask users directly. What confused you? What stopped you from signing up? What concerns do you have? What information is missing?
Surveys tell you why analytics can’t. They reveal objections, concerns, and misunderstandings that quantitative data never shows.
Method 6: User Testing
Watch real people use your site while they talk through their thought process. What catches their attention? What do they miss? Where do they hesitate?
User testing reveals problems you never knew existed because you’re too close to your own product.
Why All Six Matter
Each method reveals different problems.
Analytics shows you where people drop off but not why. Heat maps show behavior but not motivation. Surveys reveal concerns but people can’t always articulate the real issue. User testing shows real interactions but with small sample sizes.
Use multiple methods and problems become obvious. You’ll see the same issues appear across different data sources. That’s when you know you’ve found something real.
The Research Process
Start with analytics. Identify your biggest drop off points. Those are your highest value problems to solve.
Then use 2-3 other research methods to understand why people drop off at those points. Maybe analytics shows 60% of people abandon your pricing page. Heat maps reveal they’re not scrolling to see the full feature list. User testing shows they can’t find pricing information quickly. Now you have a clear problem to test.
How Much Research Is Enough?
For small tests on secondary pages: one or two research methods.
For major tests on critical pages: three to five research methods.
For complete redesigns or new features: all six methods.
The bigger the decision, the more research you need. But even quick research beats no research.
━━━━━━━━━━━━━━━━━━━━━━
💡 QUICK WIN
Pick your most important conversion page right now. Open Google or Adobe Analytics and find the drop off rate. That’s your starting point. Everything else builds from there.
━━━━━━━━━━━━━━━━━━━━━━
Coming up in Part 2:
Heuristic Analysis: how to audit any page in under an hour using a proven framework.
Reply with questions anytime.
– Atticus
P.S. Most failed tests fail because people tested the wrong thing, not because they ran the test incorrectly. Research solves that problem.