Let me clearly state that I prefer placebos over prescription drugs. The side effects don't last as long.
Maybe it's just me, but some people would rather look at a colorful graph than answer a bunch of questions. Don't get me wrong. Data visualization is an excellent way to present information. The problem is what a chart can make you feel before you know what it actually means.
Take a look at this chart:
You can't help but get excited about that jump on the 15th. It's dramatic. You must be doing something right.
That's the placebo effect of graphs and charts. The chart makes you feel better before you know whether anything actually got better.
Now look at the same data over 90 days:
Not so exciting now, is it? The spike happens every other Tuesday, right when the newsletter goes out. Buzzkill.
That doesn't mean the newsletter isn't working. It means the spike isn't new. And unless those visitors call, fill out a form, or buy something, it may not mean much.
1. The date range is too short.
A 7- or 14-day view can make any normal bump look like a breakthrough. Before you react, zoom out to 90 days or a full year.
2. There's nothing to compare it to.
"Traffic is up 40%" sounds great until you learn it was up 60% this time last year. In GA4, use the date-range comparison to compare this period with the previous period or the same period last year.
3. The chart measures attention, not customers.
Sessions, impressions, and followers are easy to chart. Calls, form submissions, and sales are what pay the bills. If your report has plenty of traffic charts and no key events, it's showing you the placebo. I wrote more about this in Why Ranking on Page One Doesn't Guarantee More Customers.
4. Nobody wrote down what happened.
We did X, and we got Y, so our marketing worked. Maybe. Or maybe a competitor's ad stopped running, a holiday hit, or bots showed up. (Bots are more common than you'd think. See Understanding Bot Traffic.) Without a record of what changed and when, you're guessing.
The fix for number four is a simple, often-overlooked feature called annotations: short notes attached to a date on your charts. They record what you changed, so you can connect cause and effect.
When I first wrote this post in 2017, annotations were hidden in the old version of Google Analytics. When Google moved everyone to GA4, they disappeared for years. They're back now, and they're also in Google Search Console.
According to Google Analytics Help:
Your annotation appears on every report with a line graph in that property, not just the one you started from. You need Analyst access or higher to create annotations, and each property can hold up to 1,000. To see all of them in one place, go to Admin → Data display → Annotations.
Google Analytics also adds its own annotations when something significant affects your data. Those can't be edited or deleted, so pay attention when you see one you didn't create.
Google added custom annotations to Search Console in November 2025. According to Search Console Help:
A few things to know: each property can have up to 200 annotations; they can't be edited (only deleted), and they're automatically deleted after 500 days. Everyone with access to the property can see them, so keep names and phone numbers out.
If it could move your numbers, write it down. For example:
Once you've been annotating for a few months, the useful questions get easier to answer:
A spike that repeats, comes from the right source, and brings in leads is real. A one-time spike with no leads is a nice graph.
Next time someone shows you a chart that's up and to the right, ask: "What happened on that date, and did it bring in customers?"
A good marketing partner can answer both parts. If all you get is another chart, read Red Flags When Your Marketing Agency Hides Your Data.
Totally nerdy blog, I know. Don't be a hater. I got certified in Google Analytics before I became a marketer.
Want help seeing what your numbers really say? We set up GA4 tracking, key events, and reports that tie your marketing to calls, leads, and customers, not just traffic. Contact Analytics That Profit to talk it through.