SEO Predictions Are Horoscopes

The quarterly business review is in eleven minutes. Your VP of Marketing has requested, quote, "the SEO forecast for Q3." You have been asked to predict the future. Not the distant future, not the abstract future, but a specific future with specific numbers that will appear on a specific slide that will be projected onto a specific wall while executives nod and pretend to understand what organic impressions means.

You open a spreadsheet. You stare at it. It stares back. Somewhere in the cells between B7 and D14, you must conjure a number. The number must be large enough to justify your salary but small enough to be plausibly achieved. The number must look scientific. The number must have decimal points. The number must, above all, exist, because the meeting is in ten minutes and the alternative to providing a number is explaining why you cannot provide a number, and that conversation is worse than simply making one up.

So you make one up.

You look at last quarter's traffic. You add fifteen percent because that sounds like growth. You subtract five percent because that sounds like realism. You round to a number that ends in zero because round numbers look deliberate. You format it in bold. You add a footnote that says "projection assumes stable algorithm environment" because that footnote will save you later when the projection is wrong and everyone has forgotten the footnote exists. You take a breath. You walk into the meeting. You say the number out loud. The executives nod. The number becomes real.

This is how SEO predictions work. This is how they have always worked. Every forecast you have ever received was produced by this exact process, or something spiritually identical to it, by someone who knew it was fiction and delivered it anyway because the alternative was unemployment.

The Anatomy of a Forecast

Let me tell you how the sausage gets made. Let me show you what happens between the client asking "what should we expect from SEO this year" and the agency responding with a PowerPoint deck full of confident arrows pointing upward.

Step one: someone opens Google Analytics. They export twelve months of data. They paste it into a spreadsheet. They calculate the average monthly traffic and the average monthly growth rate. If the growth rate is positive, they extrapolate it forward. If the growth rate is negative, they find a reason to exclude the bad months from the calculation. Seasonality, they say. Algorithm volatility. A tracking code issue that definitely existed. The number must go up. The number always goes up. The number goes up because that is what forecasts are for.

Step two: someone adds confidence intervals. Not because they calculated confidence intervals. Nobody calculated confidence intervals. They add confidence intervals because confidence intervals look statistical, and looking statistical is the entire point. The intervals are arbitrary. Plus or minus twenty percent is popular. Plus or minus fifteen percent if you want to seem more precise. The intervals exist to give you an escape hatch. If you hit the low end of your range, you were right. If you miss the range entirely, the range was too narrow, and also there was an algorithm update, and also the client didn't publish the content you recommended, and also Mercury was in retrograde.

Step three: someone adds caveats. The caveats are crucial. The caveats are where the real work happens. "Assuming consistent content velocity." "Contingent on technical recommendations being implemented." "Dependent on competitive landscape remaining stable." The caveats are an insurance policy against reality, which has a well documented tendency to diverge from spreadsheets. When the forecast fails, and it will fail, you will point at the caveats. You will say: I told you this was contingent. I told you this assumed. I told you, right there, in the footnote, on slide seven, in nine point font, that this was not a guarantee. You did tell them. They did not listen. Nobody ever listens to the caveats.

Step four: someone presents the forecast with a straight face. This is the hardest part. This is the part where you must look a human being in the eyes and tell them that you know what Google will do in six months even though Google does not know what Google will do in six months, even though Google's own engineers are debugging ranking changes in real time, even though the entire system is a black box wrapped in a mystery wrapped in a documentation page that contradicts itself. You must do this without laughing. You must do this without crying. You must do this as if you believe it, because if you do not believe it, why would they?

They believe it. They always believe it. And then Q3 arrives, and the forecast is wrong, and everyone is too polite to mention it.

The Barnum Effect Wears a Suit

In 1948, a psychologist named Bertram Forer gave his students a personality test. A week later he handed each of them a personalized assessment based on their answers. He asked them to rate how accurate the assessment was. The average rating was 4.3 out of 5. The students were astonished by how well the test had captured their essence.

The twist, and you saw this coming, and yet somehow it still lands: every student received the exact same assessment. Forer had copied it from a newsstand astrology column. "You have a tendency to be critical of yourself." "You have a great need for other people to like and admire you." "Some of your aspirations tend to be unrealistic."

The Barnum Effect, named after the circus showman who understood that people will believe almost anything if you phrase it right, is the psychological bedrock of every horoscope ever written. Make the statement vague enough to apply to everyone. Make it specific enough to feel personal. Let the reader do the work of finding meaning. They will find meaning. They always find meaning. The human brain is a pattern matching machine that runs even when there are no patterns.

SEO forecasts operate on identical machinery.

"We expect to see meaningful traffic growth in the second half of the year as our content strategy matures." This means nothing. This is horoscope. This applies to every SEO engagement that has ever existed. And yet the client nods, because meaningful sounds measured, because second half of the year sounds specific, because content strategy matures sounds like a thing that happens, like wine aging or cheese ripening or software becoming enterprise-grade.

"Rankings volatility is likely in the near term, but we anticipate stabilization as domain authority builds." Translation: things might get worse before they get better, or they might get better before they get worse, or they might stay the same, and in six months we will point at whichever outcome occurred and call it stabilization. This is not prediction. This is cold reading. This is a psychic at a county fair telling you that someone whose name starts with M has unfinished business with you.

"We're cautiously optimistic about hitting our targets, assuming the recommendations are implemented and the competitive landscape remains stable." This sentence contains zero information. It is a perfectly balanced equation where every variable cancels out. Cautiously optimistic: could go either way. Assuming recommendations are implemented: if it fails, it's your fault for not implementing. Competitive landscape remains stable: the competitive landscape has never once remained stable in the history of search engines. The caveats devour the prediction. The prediction was never a prediction. The prediction was a permission structure for future blame allocation.

The Three to Six Month Hedge

How long until we see results?

This is the question. This is the only question. Every client asks it. Every stakeholder asks it. Every VP who just approved a six figure SEO budget asks it while staring at you with eyes that say, I need to report something to my board in ninety days and it cannot be "we are still waiting."

The answer, always, forever, across all agencies and all consultants and all in-house SEOs from the beginning of search engine optimization unto the heat death of the universe, is: three to six months.

Three to six months is the horoscope answer. Three to six months is the number that buys you time without committing to anything. Three to six months is short enough to sound actionable and long enough that everyone will have forgotten the original question by the time the deadline arrives. Three to six months is the SEO equivalent of "it depends," weaponized into a timeline.

Here is why three to six months is always the answer:

If you say results in one month, you are a liar, and you will be caught.

If you say results in one year, you will not be hired, because nobody has one year of patience, because every marketing budget operates on quarterly cycles, because the CMO's job security extends approximately nine months and they need a win before then.

So you say three to six months. And if the results arrive in three months, you are a genius. And if the results arrive in six months, you met expectations. And if the results arrive in nine months, well, there was an algorithm update, and the competitive landscape shifted, and the content took longer to index than anticipated, and anyway nine months is basically six months plus seasonality.

And if the results never arrive? You will not be there. The contract will have ended. The internal champion will have left for another company. The strategy will have pivoted to paid acquisition. The question "did SEO work" will dissolve into the organizational amnesia that claims all inconvenient measurements. Nobody ever goes back and checks. Nobody ever pulls up the original forecast and says, wait, you said 40% traffic growth and we got 3%, what happened. That conversation does not occur. That conversation has never occurred. That conversation would require someone to admit that they believed a horoscope.

The In-House SEO's Existential Nightmare

I have been hard on agencies. I have been hard on consultants. Let me now be hard on the people who deserve the most sympathy, which is to say the in-house SEOs, the ones who cannot simply end the contract and walk away, the ones who must sit in meetings every week with executives who think rankings are a dial you turn.

The in-house SEO is asked to predict the future more than anyone. The in-house SEO is asked to produce forecasts on demand, to conjure traffic projections from nothing, to explain why the hockey stick graph from last quarter's deck has become a plateau or, god forbid, a ski slope in the wrong direction. The in-house SEO reports to people who do not understand what SEO is, who think it is a button you press or a vendor you hire or a cost center that should produce measurable ROI within one fiscal quarter.

"What's the SEO opportunity for this new product launch?"

This question arrives in your inbox on a Tuesday afternoon. The product launches in six weeks. You have been given no keyword research, no competitive analysis, no information about the product beyond a two paragraph brief written by a PM who thinks SEO is something you sprinkle on content like salt. You must respond by end of day because there is a leadership meeting tomorrow and they want to include the SEO opportunity in the slide deck.

So you open a keyword tool. You type in the product category. You export the search volume data. You add it up. You subtract some percentage because not all of that volume is actually capturable. You add some percentage because long tail queries exist and the tools undercount them. You arrive at a number. The number is meaningless. The number is an educated guess wrapped in a spreadsheet wrapped in false precision. But the number exists, and existence is all that was required.

The number goes into the deck. The leadership team nods. Someone asks about timeline. You say three to six months. Someone asks about confidence level. You say medium to high, contingent on content and technical implementation. Someone asks what medium to high means as a percentage. You say seventy percent, because seventy percent sounds more confident than sixty percent and less arrogant than eighty percent. Someone writes down seventy percent as if it were a real measurement of a real thing.

Six months later the product has launched and the traffic is 12% of projection and nobody asks why. Nobody asks why because asking why would require revisiting the original projection, and revisiting the original projection would reveal that the projection was a guess, and revealing that the projection was a guess would undermine all the other projections currently circulating through the organization, and those projections are load-bearing, those projections are holding up roadmaps and budgets and executive bonuses, and if you pull out one brick the whole wall might come down.

So nobody asks. And you make another projection. And the cycle continues.

Why We Do This

Let me extend good faith. Let me be the fairest man alive for one paragraph. Let me acknowledge the human reality of why predictions happen.

Predictions happen because people need to plan. Budgets are allocated annually. Headcount is justified quarterly. Marketing teams must commit to goals before they know whether those goals are achievable, because that is how organizations work, because that is how capitalism works, because the alternative is chaos and paralysis and no one ever doing anything because the future is uncertain and will always be uncertain.

The CEO asks the CMO for a growth target. The CMO asks the VP of Marketing for a channel breakdown. The VP asks the Director of Acquisition for an SEO number. The Director asks the SEO Manager for a forecast. Each layer adds pressure. Each layer demands specificity. By the time the question reaches the person who actually does SEO, it has transformed from "what might happen" into "what will happen," and the person who actually does SEO must answer as if the future is knowable because their job depends on acting as if the future is knowable.

I understand this. I have lived this. I have been the person in the meeting making up numbers because the alternative was admitting that the entire planning process is built on sand.

But understanding why we do it does not make it less absurd. Understanding the incentive structure does not make the forecasts less fictional. The emperor is naked and we all know the emperor is naked and we have built an entire industry around pretending to admire the stitching on his invisible clothes.

The Algorithm Does Not Care About Your Forecast

Here is the thing that makes SEO predictions categorically different from, say, sales forecasts or manufacturing projections or any other kind of business prediction that has at least some grounding in controllable variables:

You do not control Google.

When a sales team forecasts revenue, they are forecasting something that depends on their own effort. They can hire more reps, make more calls, adjust pricing, offer discounts. The inputs are, to some extent, within their control. When SEO forecasts traffic, we are forecasting something that depends on the whims of a company in Mountain View that does not know we exist, that changes its algorithm thousands of times per year, that occasionally releases an update that obliterates entire industries overnight with no warning and no explanation.

You can do everything right. You can publish the content. You can build the links. You can fix the technical issues. You can follow every best practice ever documented. And then an update rolls out and your traffic drops 40% and nobody at Google will tell you why, and your forecast becomes confetti, and the executive who approved your budget is now looking at you as if you personally betrayed them.

Or, alternatively: you can do everything wrong. You can ignore the recommendations. You can publish thin content. You can acquire spammy links. And then an update rolls out and your competitors get crushed and you gain traffic by default, and suddenly you are a genius, and your forecast was exceeded, and no one asks too many questions about why.

This is the reality. The signal to noise ratio between effort and outcome is so low that prediction becomes essentially random. You are forecasting weather patterns in a system where the sun might randomly explode. You are projecting stock prices in a market where the SEC might randomly criminalize trading. You are, and I cannot stress this enough, completely at the mercy of an entity that owes you nothing and will never acknowledge your existence.

And yet we forecast. And yet we predict. Because the alternative is honesty, and honesty sounds like this:

"I don't know what will happen. I have no idea. I can tell you what we will do, but I cannot tell you how Google will respond, because Google has never once told me how Google will respond, because Google changes its mind constantly, because Google is not a coherent entity with consistent goals but rather a collection of teams running experiments that sometimes contradict each other. The best I can offer you is educated guessing. The best case is that our educated guessing is slightly less wrong than random guessing. Would you like to proceed on that basis?"

Nobody says this. Nobody can say this. The meeting would end. The contract would evaporate. The budget would disappear. So instead we say: three to six months. And everyone pretends to believe it. And the ritual continues.

An SEO forecast is a horoscope wearing a spreadsheet's clothes. The stars are Google's servers. The astrologer is your agency. The only difference is that astrologers occasionally admit they're interpreting signs rather than stating facts.

What to Do Instead

I have spent two thousand words telling you that predictions are fake. Let me spend two hundred telling you what is not fake.

Scenarios, not forecasts. Instead of saying "we will get 50,000 visits," say "here are three scenarios: conservative, moderate, aggressive. Here is what would need to be true for each one. Here are the signals we will watch to know which scenario we are in." This is still prediction, but it is honest prediction. It admits uncertainty rather than cosplaying certainty.

Leading indicators, not trailing outcomes. Instead of forecasting traffic, forecast what you can control. How many pages will you publish? How many links will you build? How many technical issues will you fix? Forecast the inputs. Let the outputs be what they will be. At least you will know whether you did the work.

Retrospectives, not projections. The best predictor of what will happen is what has already happened. Instead of asking "what will our traffic be in Q3," ask "what did our traffic do last Q3, and what did we do differently then, and what can we learn from the variance." Look backward to see forward. It is not perfect but it is less fake.

Ranges, not points. If you must give a number, give a range so wide that it cannot be wrong. "Between 20,000 and 80,000 visits" is useless for planning but honest about uncertainty. If the stakeholder pushes back and says they need a specific number, you now have an opportunity to have a real conversation about why specific numbers are impossible. Take that opportunity. It might be the first honest conversation about SEO anyone in that room has ever had.

Admissions, not hedges. The difference between a hedge and an admission is courage. A hedge says: "assuming stable algorithm environment." An admission says: "I do not know what Google will do and neither does anyone else, including Google." The hedge lets everyone pretend. The admission lets everyone breathe.

I know. None of this is practical. The meeting is in six minutes and the VP needs a number and the number must exist and you cannot walk in there and deliver a philosophy lecture about epistemological humility in the face of algorithmic chaos. I know.

But maybe, next time, you can at least laugh quietly to yourself as you format the cells in bold. Maybe you can acknowledge, internally, that you are participating in a ritual that has more in common with astrology than analytics. Maybe that acknowledgment, even if never spoken aloud, is a kind of freedom.

The forecast is fake. The forecast has always been fake. We are all just guessing, dressed in the costume of science, performing confidence for an audience that desperately wants to believe.

Mercury is in retrograde. The algorithm is volatile. Your traffic projection is written in the stars.

Three to six months. Give or take.