I have watched more business owners panic over a dipping traffic line than I can count, and almost none of them stopped to ask whether the money had actually moved. That gap between the graph and the bank account is the whole game, and two recent conversations on the Unscripted SEO podcast sharpened how I talk about it.
Traffic was never the thing you were buying
Ernesto Ortiz of Structural SEO came into search from an engineering background, running single-variable tests on his own domains before he trusted anything anyone told him. So when he says traffic is a measurement of an intermediate step, he means it the way an engineer means it: it is a sensor reading, not the output you actually care about.
Traffic is a proxy. It always has been. If you see the graphs looking bad, ask: are you actually losing sales or are you just losing traffic? People still need to go and buy your thing. It’s irrelevant if they read your article or if the AI gave them the answer.
— Ernesto Ortiz
For an e-commerce store, a traffic decline that does not touch revenue per session or total sales is not a crisis. It often means AI overviews are absorbing top-of-funnel queries that were never going to convert anyway. The people who were going to buy still arrive and still check out. The line on the chart fell; the register did not. That is a very different emergency than the one most dashboards scream about.
Publishers are the exception, and it is worth naming why. Their model needs the click so the ad can render. If the answer shows up before the visit, the visit never happens and neither does the impression. For them the business model has to change, not just the keyword plan. Knowing which side of that line your client sits on is step one, and it is the kind of thing I keep hammering in my search fundamentals work with newer practitioners.
Algorithm updates only hurt the incomplete
Here is the framing from Ernesto that I have started stealing outright. Updates are not new rules dropped from the sky. They are the existing dials being re-weighted.
Algorithm updates are just turning the dials on factor importance. If you go down, it’s because you’re incomplete — you optimized for the factors that were currently tuned up, and when they changed the dials, you didn’t have the rest. If you have a more complete situation, you either go up or nothing happens.
— Ernesto Ortiz
If you leaned your whole ranking on one heavily weighted signal, a re-weighting exposes that you skipped the rest. Build a complete picture across content, links, technical health, and user experience, and an update has nothing to catch. Going down is information about what is missing, not a verdict on your worth. This is also why I get nervous watching the manipulation arms race I wrote about in what Timothy Malmros taught me about ranking spam — thin tactics work right up until a dial turns.
Ernesto’s risk filter for any aggressive shortcut is one I now ask out loud with clients: would you bet the future of your business on this tactic? Most legitimate operators say no, and that no is usually correct.
Measure the pages where money changes hands
If traffic is the wrong headline metric, what is the right one? For e-commerce, Ernesto’s primary KPI is non-brand traffic to category and product pages. Not total organic. Not blog sessions. Not branded search, which mostly measures your other marketing. Blog posts exist to push internal-link equity into the pages that can actually take a credit card; their traffic is instrumental, not the goal.
Pair non-brand traffic on money pages with branded-search trends and you can separate what SEO contributed from what the brand was already going to earn. Report revenue, yes, but also report the pure SEO metric that proves you moved the right needle. That discipline is the same budget-flipping logic behind what Greg Digneo taught me about flipping the SEO budget: spend against the outcome, not the vanity number.
Being seen still counts when you write for humans
Julia Bocchese runs Julia Renee Consulting for creative and holistic small businesses, and she teaches SEO at the college level without a textbook because the field moves too fast for one. She makes the revenue point from the conversion side instead of the analytics side.
I keep seeing clients focusing their copy more on writing for the bots crawling their website rather than writing for the humans that are actually going to be hiring them and giving them money.
— Julia Bocchese
The bots do not have a wallet. A page can rank and still fail to convince the person who could pay you, which is exactly the trap of optimizing the proxy instead of the sale. Julia also pushes a wider read on reporting: when clicks dip, impressions and engagement quality still carry value, because being seen by the right searcher is a step toward the eventual purchase. She is bullish on Pinterest as a search engine for the same reason — intent-driven discovery that sends qualified humans, not vanity reach. It is practical, budget-aware advice I would hand to any owner listening to Unscripted Small Business.
Build the habit, not just the dashboard
The fix is mostly a habit. Before you react to a traffic graph, pull the revenue line next to it. If sales held, breathe and keep building completeness. If sales fell, then you have a real problem worth chasing. I lean on the conversion and tracking tooling over at SEO Arcade to keep that comparison honest, and I work through this exact reframe with the entrepreneurs at the Digital Christian Collaborative here in Cookeville.
After twenty years in this work, the lesson keeps repeating: traffic is the sensor, revenue is the engine. Watch the engine. Build complete signals so the next update has nothing to grab, write for the human holding the credit card, and the scary graphs stop running your week.