---
title: "The Market Analysis That Made Us Cancel an Ad Campaign Before It Launched"
url: "https://marketermagazine.co/insight/the-market-analysis-that-made-us-cancel-an-ad-campaign-before-it-launched/"
author: "Nassira Sennoune"
published: "2026-09-25"
updated: "2026-09-25"
---

# The Market Analysis That Made Us Cancel an Ad Campaign Before It Launched

**TLDR:** Our brokerage had a paid search campaign ready to go on the broad terms every luxury property agency in Marrakech bids on. A two-week market analysis, done by buyer origin and query language rather than by keyword volume, showed those terms belonged to portals we could never outbid, and that the questions our actual buyers typed had almost no competition at all. We cancelled the campaign, rebuilt the site around the questions, and kept one small ad campaign that only talks to people who have already been on a property page.

The campaign was built. The ad groups were named, the budget was approved, the landing page had been signed off by the head of sales, and the launch was set for a Monday. I work in-house on the digital side of a luxury real estate brokerage in Marrakech, and my job that week was supposed to be pressing the button.

Instead I asked for two weeks. Not because the campaign looked wrong. Because nobody, including me, could answer a simple question about it: who exactly was going to type these words, and from where?

### **The Words Everyone Buys**

The campaign targeted the obvious phrases. Luxury villas in the city. Property for sale in the city. Real estate agency plus the city name. They had the volume, they were what the previous agency had recommended, and they were what every competitor's ads showed when you searched them.

That last part should have been the warning. When I ran the searches myself, the paid results and the top organic results were the same handful of names: international listing portals, aggregators with tens of thousands of pages, and two or three large agencies with a decade of history. A brokerage with sixty pages on its website was not going to become the best answer for "property for sale in Marrakech" by paying for clicks, and if it did rank, it would rank eleventh.

Ahrefs describes this trap well in its guide to [search intent](https://ahrefs.com/blog/search-intent/): the volume on a broad query tells you how many people typed it, not whether your page can be the thing they wanted when they did. **On a broad property term, the thing they wanted was a portal.** We were about to pay to send buyers to a page that would lose the comparison.

### **Who Actually Buys From Us**

So the analysis started somewhere else: with the closed deals of the previous two years, not with a keyword tool.

Our buyers mostly do not live in Morocco. They come from France, Belgium and Switzerland, from the Gulf, and increasingly from West Africa. They search in French and in English, sometimes in both in the same session. And when I went back through the first emails and first calls, the questions were almost never "villas for sale." They were specific and slightly anxious:

- Can a foreigner own property outright here, and what does the notary actually do?
- What is the real difference between a riad in the medina and a villa in the Palmeraie for someone who will visit six weeks a year?
- How far is this neighborhood from the airport, and from an international school?
- What happens to the money when I sell, and in which currency?

Then I put those questions into the keyword tools. Each one had a fraction of the volume of the broad terms. Together they were a different story. And the pages answering them were thin, old, or written by lawyers for other lawyers. _For a buyer, there was almost nothing to read._

That gap is the whole finding. A market analysis by buyer origin and by question, instead of by volume, showed us a market with two halves: a loud half owned by portals, and a quiet half nobody had bothered to serve.

### **What the Site Audit Added**

The second week, I audited our own site against that map, and it was humbling. Of the sixty pages we had, twenty-six were effectively invisible to Google: duplicated neighborhood pages, property-type pages that repeated each other, and area guides that said nothing a portal did not say better. The keyword mapping that came out of the analysis suggested several hundred candidate pages worth building over time, against sixty-three that existed, and most of the existing ones were competing with each other for the same handful of queries.

Google's own guidance on [creating helpful content](https://developers.google.com/search/docs/fundamentals/creating-helpful-content) asks a plain question: would someone feel they had learned enough about the topic to reach their goal? For a buyer in Brussels wondering whether they can legally own a riad, our pages did not pass that test. An ad campaign sending that buyer to those pages would have paid for the disappointment.

We cancelled the launch that Friday.

### **What We Built Instead**

The budget did not disappear. It was redirected, and most of it stopped being media spend.

- **One page per real buyer question.** Ownership for foreigners, the notary process, currency and repatriation, the honest comparison between neighborhoods. Written in the language the question was asked in, with the specifics a buyer wants rather than the reassurance an agency prefers to give.
- **Consolidation before creation.** The twenty-six invisible pages were merged into the ones that had a reason to exist. Fewer pages, each with one job and its own inventory.
- **Inventory on every area page.** Real available properties with real prices, because a page about a neighborhood without listings is a blog post, and a buyer knows the difference.
- **One small ad campaign, kept.** Retargeting only, shown to people who had already spent time on a property page, with the property they looked at and one line about the next step. It costs a fraction of the cancelled campaign and it talks to people who have already told us what they want.

The innovation, if there was one, was not clever. It was refusing to buy attention for a question we could not answer better than a portal, and spending the effort on the questions we could answer better than anyone.

### **The Signal Nobody Was Measuring Yet**

One more thing came out of the analysis, almost by accident. While checking how buyers phrase their questions, I started asking AI assistants the same questions a buyer would ask: which brokerage to use in the city, whether a foreigner can buy, which neighborhood suits a family. The answers were built from exactly the kind of specific, question-shaped pages we had just decided to write, and they quoted the sources by name.

That changed how the pages were written. Each one now opens with a direct answer in the first paragraph, states the fact and the exception, and names the source of the rule. It is the format a buyer scans on a phone, and it happens to be the format an assistant can quote. We treat it as a second market, and I now ask every new buyer on the first call whether they used an assistant before they searched.

### **What Changed on the Sales Floor**

The head of sales did not care about invisible pages. What he noticed, a few months in, was that first calls got shorter. Buyers arriving from the question pages had already read the ownership explainer and the neighborhood comparison, so the call started at the property instead of at the legal basics. Enquiries came in with the neighborhood already chosen. And the retargeting ads, the one campaign we kept, produced viewing requests from people the brokerage had never spoken to but who had clearly done their reading on [the brokerage's website](https://originnproperties.com/) first, on the property pages and the area guides that the analysis had told us to rebuild.

I will not pretend the cancelled campaign would have been a disaster. It would have produced clicks, a report, and a reason to run it again. What it would not have produced is a site that answers the questions our buyers actually ask, and that is the asset that keeps earning after the budget is spent.

### **Run the Analysis Before the Campaign**

If you are about to launch a campaign on your category's broadest terms, spend two weeks on this first:

- Pull your last two years of customers and write down where they came from and what they asked first. Do this before opening a keyword tool.
- Search your broad terms yourself and look at who already owns the page. If you cannot honestly be the best answer, the volume is not yours.
- Map the specific questions your customers asked to pages on your site. Count the questions that have no page. That list is your content plan and, often, your real ad plan.
- Keep one campaign that only talks to people who have already shown intent on your site. It is the cheapest ad you will ever run and the easiest to defend.

The uncomfortable part is telling the people who approved the campaign that the market analysis says not to launch it. In our case it helped that the analysis was two weeks of work, not two months, and that it ended with a plan rather than a warning.

---

Nassira Sennoune is the SEO consultant at Originn Properties, a luxury real estate brokerage based in Marrakech, working with buyers and owners across Morocco and the Gulf. If you are weighing a campaign on your market's broadest terms and want to see how the brokerage answers the questions its buyers actually ask, you can find the team at [Originn Properties](https://originnproperties.com/).