If you've ever gotten excited about fixing your analytics setup and then had to sit across from your leadership team explaining why the numbers look worse than last quarter, you know the particular discomfort this episode is about. That conversation is coming for you sooner or later, and how you frame it when it does will matter.
Sarah and I walked through the pattern we repeatedly see when companies undergo a data cleanup, whether it's triggered by an agency switch, new leadership, an M&A event, or something as fundamental as the GA4 migration. You'll come away with a clearer picture of what "cleaner data" actually means versus "fixed data," how to identify the two data inflators Jay finds on nearly every new client onboard, and how to guide a nervous executive through a data reset without losing their confidence in the process.
This post is based on Episode 61 of Revenue Rewired | EP 61: Why Good Data Feels Bad at First.
If you'd rather listen than read, find the full episode on Apple Podcasts, YouTube, Spotify, or Amazon. It's worth your time.
When the Numbers Drop After You Fix Things, What Does That Mean?
Thirty years ago, I was part of a firm that a client hired to clean up their entire funnel, tools, and analytics. We were proud of the work. When they asked for a year-over-year report, the numbers showed a significant decline. The client was worried about their job and the investment they'd made. What I had to explain, over multiple conversations, was that the old data was inflated, and we were finally looking at accurate numbers. The reports had looked right. The decisions built on them had been wrong.
That experience changed how I handle data conversations with every client since. Now, before we do anything else, I tell clients upfront that we work from honest data, and if what they've been reporting isn't accurate, we're going to surface that. The discomfort isn't a sign that something went wrong. It's what it feels like to finally see a real starting point.
What makes it hard for the people in the room isn't the drop itself. It's that the numbers going down look like failure to anyone who isn't close to the work. That's the communication challenge, and it doesn't go away just because you can explain the methodology.
What Did the GA4 Migration Actually Do to Your Year-Over-Year Numbers?
Sarah called it a marketing earthquake, and that's about right. The switch from Universal Analytics to GA4 wasn't just a platform change. It rewrote the baseline for year-over-year comparisons. Bounce rate as a metric disappeared. Engagement rate replaced it, measuring something different enough that layering old numbers on top of new ones produces comparisons that are hard to defend in a leadership meeting.
We also had clients running Google Ads and Google Analytics side by side, both Google products, and getting meaningfully different visitor counts for the same campaigns. One client's ad platform showed 23,000 visits for the month. Their analytics platform showed 21,000. Not because something was broken, but because the two platforms attribute and count differently. Explaining that gap to a CMO who just wants to know whether the campaign worked is its own challenge, and you don't get to skip it.
The single-source-of-truth conversation is about this specifically: pick one platform, use it consistently, and build your comparisons within that consistent context. Your data isn't going to be perfect. But it can tell a reliable story if you're pulling from the same place the same way every time, and if you're explicit about what changed when you do make a system change.
What's Likely Inflating Your Traffic Right Now?
When StringCan onboards a new client, we run through a 30-plus-point checklist to assess the reporting foundation. Two issues show up with enough regularity that I want to name them specifically, and neither one is bots, though those exist too.
The first is internal employee traffic. If your team logs into your site daily and your company's IP addresses aren't excluded from analytics, every one of those sessions counts. If you have remote employees, their home IPs need to be excluded as well. This is a five-minute check that a lot of companies skip entirely.
The second is partner login traffic. Many B2B companies have a partner or dealer network that accesses a client portal through the company's main website homepage, clicking a login button before they ever engage with any marketing content. Those sessions register as real visits with near-zero engagement. The traffic numbers look impressive. The story they're telling is wrong. Both of these inflate your counts and tank your engagement rate, and together they're the data inflation issue I find on almost every new client audit.
Why Is It So Hard to Let Go of the Numbers You're Used To?
Marketers are an emotional group, and I say that having been one for thirty years. We rely on metrics to justify the investment our organizations are making in us, and when those numbers shift, even for good reasons, it can feel like losing the evidence. Nobody wants to walk into a leadership meeting and explain that the campaign they were celebrating last quarter was actually built on inflated data. That's a hard conversation regardless of how reasonable the explanation is.
The 54-page PowerPoint is the extreme version of this dynamic, but I watched it happen in real life. A talented analytical colleague put one together for the CMO of a large real estate firm. He was proud of it. I sat through the presentation, and by slide seven, I didn't understand what he was saying anymore, and I like data. I kept picturing the CMO, a creative marketer who wasn't analytical, and thinking she would have been lost by slide two. The problem wasn't the data. It was that no one had asked "so what?" at every step before building the deck.
That's the reframe that changes reporting conversations: the question isn't "what does this report show?" It's "what are we trying to answer, and does this number tell us anything about that?" If a data point doesn't connect to a decision someone is actually going to make, it probably doesn't belong in front of leadership. Clean data and clear communication are two separate skills, and you need both.
FAQ
Q: How do I explain to my CEO that our traffic numbers were inflated before, without making the marketing team look incompetent?
A: Set expectations before the numbers change, not after. Tell leadership upfront that you're going through a reporting cleanup and that accurate data almost always shows lower initial numbers than what the team was used to seeing. The framing matters: this is a reset to a reliable baseline, not a performance decline. Coming in with the specific reasons for the inflation, internal traffic, partner logins, and platform discrepancies gives them something concrete to hold onto instead of just a drop.
Q: Is there a quick way to know whether we have a data inflation problem before doing a full audit?
A: Two checks take about three to five minutes each. First, confirm that your company's internal IP addresses are excluded from your analytics platform, including any remote employee home IPs if you have them. Second, look at whether your partner or vendor network accesses a client portal through your main website homepage. If they do and there's no filter in place, those sessions are registering as real website traffic with minimal engagement. If either of those is missing, you likely have a meaningful inflation problem in your current data.
Q: What does "single source of truth" actually look like for a company with ten or more tools in its tech stack?
A: It means picking one reporting platform and using it consistently for your comparisons, with the full understanding that it won't be perfectly accurate. A typical mid-market B2B company has ten to fifteen different tech stack items that touch attribution in some way. You're not going to reconcile all of them. What you can do is commit to one consistent system so that week-over-week and year-over-year comparisons are at least internally coherent, and you're not explaining methodology shifts every time someone asks how last month compared to the one before it.
Q: Our numbers dropped significantly after the GA4 migration. How do we know what's real versus what's a measurement artifact?
A: With real caution about year-over-year comparisons across the migration period. The methodology shift was significant enough that layering GA4 data on top of Universal Analytics data creates gaps that can look like performance problems when they're actually just a change in how visits, sessions, and engagement are being counted. If you're doing cross-period comparisons, you need to explicitly call out the measurement difference, or you're going to confuse your leadership team and potentially make channel decisions based on a drop that doesn't mean what it looks like.
Q: At what point should we stop fixing the foundation and start adding new campaigns?
A: After you've stabilized what you already have. Adding new campaigns on top of a broken reporting foundation just creates more variables to explain later, because now you can't tell whether a performance change came from the campaign or from a data issue that already existed. Get the exclusions right, establish a consistent attribution logic, and confirm your single source for comparison before you build on top of it. That sequence matters.
Ready to Stop Making Decisions From Numbers You Can't Trust?
At StringCan, the first thing we do with a new client isn't build a strategy. It's audit the reporting foundation, because you can't make smart decisions about budget, channel mix, or campaign performance if the numbers you're working from are inflated or built across two different measurement systems. We call this pattern the Metrics Misfire in our Revenue Rewired book, and it's one of the more common revenue leaks we find in owner-led B2B companies doing $10M to $50M in revenue.
If you want to hear the full conversation, including the story behind the 54-slide dashboard and what Jay's onboarding checklist actually catches, listen to Episode 61 here. And if you're not sure whether your reporting has a Metrics Misfire, the Revenue Leak Assessment at stringcaninteractive.com/revenue-leak-assessment is a useful starting point.
