# How to market a specialty coffee store in Saudi Arabia, and know the ads pay

Published November 18, 2026 · Updated November 18, 2026

<https://mqudah.com/en/insights/coffee-store-marketing/>

A bag of beans runs out in weeks, and a happy customer comes back. So judge a coffee ad by the customer, not the order, as long as your sales sheet proves they return. Inside: an eight-row diagnostic table.

## The short answer

Don't judge a coffee ad by the first order alone. A bag runs out within weeks, and a happy customer orders again. So work out what the whole customer leaves you. But count the second order only if you see it in your sales sheet. Start with the search terms report, because in my view «كوفي», the word 'coffee' in Arabic letters, mostly brings people looking for a café nearby, not beans.

## The bottom line

- My view: in coffee the first order can look like a loss, because the bag runs out and a happy customer comes back. So judge the ad by the customer, not the order.
- Count a second order only if your sales sheet shows it, over a period you measured yourself. Never pay for a click on a return you're only hoping for.
- «كوفي» (coffee) is a huge, cheap search, but my reading is that most people who type it want a café or a drinks menu. Fix it with negative keywords and match type, not a higher bid.
- Show the roast date, origin, roast level and grind on the product page. That's what people buying beans online check before they pay.
- Make the second order easy, with a WhatsApp reminder before the bag runs out or a subscription offered after the first order.

Plenty of store owners live by one rule. If the first order doesn't cover the ad, the ad is losing money. In perfume and abayas, I think that rule mostly holds. In coffee, it's half right.

Why half? A bag of coffee runs out in a few weeks, and a customer who liked it comes back. So the first order can look like a loss while the customer makes you money. But the other half of the rule still stands. A second order counts only once you see it in your sales sheet.

I'd judge a coffee ad by the customer, not the order. A click you pay for on a return you hope for, but haven't measured, is a loss you've postponed.

## How do you know your coffee store ads aren't paying?

You know when a new customer costs you more than they leave you, over a period you've measured. I start with the sales sheet, then open the ad account:

- **Sales sheet on Salla or Zid** (two of the main Saudi store platforms): most customers from ads ordered once and never came back.
- **Order log:** orders placed well above orders delivered, mostly cash-on-delivery parcels refused at the door.
- **Search terms in Google Ads:** phrases with «قريب» (near) or «منيو» (menu), or cafés and roasters you don't sell.
- **Average cost per click:** above the ceiling you worked out per customer.
- **Product page and WhatsApp:** many visits, few add-to-carts, and repeated questions about roast date or grind.

Two of these, and the problem is rarely the ad's image. It's the sum.

## Why does this happen with coffee in particular?

Because coffee gets reordered, and because coffee searches mix buyers with people who want a café. Here are three phrases from my Keyword Planner export. The bid column estimates what advertisers pay to reach the top of the page. It isn't what you'll pay.

| Search phrase | Monthly searches | Top-of-page bid (SAR) |
|---|---|---|
| «كوفي» ("coffee" in Arabic letters) | 100k–1M | 0.35–3.00 |
| «قهوه مختصه» (specialty coffee) | 10k–100k | 0.60–2.72 |
| «متجر قهوه» (coffee store) | 1k–10k | 0.34–1.56 |

Figures from Keyword Planner, Saudi Arabia, November 2026.

**The big cheap word brings café-seekers.** «كوفي» gets far more searches than the other two, and Google rates its competition as low. But my reading is that most people typing it want a café nearby, a drinks menu or a name for their coffee shop. Few want beans. So I don't treat it as a bargain. I handle it with negative keywords and match type.

**Name searches go to known names.** My sense is that much of the coffee search carries the name of a roastery or chain people already know. A specialty coffee store nobody knows yet won't win those searches. Their size still tells you where you stand today.

**The first order is small. The profit is in the return.** One bag leaves a thin margin after shipping and payment fees. A click ceiling built on it alone can fall below most bids, as in the hypothetical example below. But a happy customer comes back when the bag runs out. That's where coffee differs from perfume and dates. In [dates and honey stores](https://mqudah.com/en/insights/dates-honey-store-marketing/), demand bunches into a season. Coffee gets reordered every few weeks, all year.

**Buyers judge freshness from the page.** They can't smell the beans before paying. So they look for the roast date, origin, roast level and grind. Without those, they hesitate, ask on WhatsApp, or buy from a store that shows them.

**Seasons bring buyers who may not return.** In Ramadan and on White Friday (the Gulf's Black Friday), some people buy coffee as a gift. A gift buyer may never reorder for themselves. So I wouldn't give seasonal buyers your regular customers' return rate. Plan both with [Ramadan campaign planning](https://mqudah.com/en/insights/ramadan-campaign-planning/) and [preparing for White Friday](https://mqudah.com/en/insights/white-friday-preparation/).

## How do you work out what a click can cost a coffee store?

Work it out twice, once per order and once per customer. Then compare both with what you pay. I always read the sales sheet before the campaign report, because the verdict on a click starts with the margin.

1. **Clean up the search terms first.** Sort the search terms report in Google Ads by cost. Look for café intent, jobs, recipes and names you don't sell. If a phrase has spent more than one order leaves you and sold nothing, put it on the decision list.
2. **Work out the click ceiling on the first order.** That's the most you can pay for one click. Take what a delivered order leaves after VAT, beans, roasting, packaging, shipping and payment fees. Multiply it by the share of visitors who buy, your conversion rate. The [campaign profit calculator](https://mqudah.com/en/tools/campaign-profit-calculator/) does the sum and deducts refusals and returns too.
3. **Then work it out per customer, from the sheet, not from hope.** Take the customers who first ordered in one month. Follow their delivered orders for six months. Add up what those orders left you, and divide by the number of customers. That's what one customer leaves you in that period.

**Hypothetical example, round numbers.** A bag of beans sells for SAR 80 and leaves you SAR 30 after all costs. One visitor in 100 buys. Now assume your sheet shows the average customer orders three times in six months.

| Hypothetical example | First order only | Customer over 6 months |
|---|---|---|
| What you keep | SAR 30 | SAR 90 from 3 orders |
| Visitors who buy | 1% | 1% |
| Most you can pay per click | SAR 0.30 | SAR 0.90 |

The first ceiling is below the bottom of the «قهوه مختصه» range in the table above. Judged on the first order, that phrase loses money. The second ceiling falls inside the range, near its low end. But the three orders are an assumption. If your sheet says customers order once, your ceiling is the first one.

4. **Separate new customers from returning ones.** Add a column to the sales sheet that says whether this is the customer's first order. The ad answers for new customers. My reading is that a campaign whose orders come mostly from existing customers pays for returns that would have happened anyway. The full method, with the payback period, is in [customer acquisition cost and lifetime value](https://mqudah.com/en/insights/cac-and-ltv/).

## Which row looks like your account this week?

Pick the symptom closest to your reports. Do what the last column says before changing anything else.

| What you see | Likely cause | Where to check | What to do |
|---|---|---|---|
| Many cheap clicks on «كوفي», few orders | Searchers want a café, not beans | Search terms report | Add «قريب» (near), «منيو» (menu) and similar as negatives. Tighten match type |
| Cafés or roasters you don't sell in search terms | Broad match catches name searches | Search terms report | Add those names as negatives |
| «وظائف» (jobs), «طريقة عمل» (how to make) or «دورة» (course) in search terms | Job seekers, recipe hunters, learners | Search terms report | Add them as negatives |
| First order loses in every campaign | Ceiling set on one order | Calculator and sales sheet | Work out what a customer leaves over a measured period |
| Customers from ads don't come back | No easy path to a second order | Second orders in the sales sheet | Add a WhatsApp reminder or a subscription |
| Most campaign orders come from existing customers | The ad pays for returns that would happen anyway | "First order?" column in the sheet | Judge the campaign on new customers only |
| Many page visits, few add-to-carts | Freshness details missing | Product page and WhatsApp questions | Show roast date, origin, roast level and grind |
| Many orders placed, fewer delivered | Refusals at the door | Order log and courier reports | Confirm each order before it leaves the warehouse |

## What do you fix first, and what can wait?

I've ordered the steps by cost, cheapest first. The budget is the last thing you touch.

**Negative keywords first.** These are words you add so your ad doesn't show on searches you don't want. Start with café intent, such as «قريب» (near), «منيو» (menu), «كافيه» and «مقهى» (both mean café). Then jobs and learning, such as «وظائف», «طريقة عمل» and «دورة». Then roasters and chains you don't sell. A negative keyword doesn't cover its other spellings, so add «قريبة», «كافي» and «كوفي شوب» as people type them. Build the list with [negative keywords for Arabic search](https://mqudah.com/en/insights/negative-keywords-arabic-search/).

**Then narrower intent.** Move budget from «كوفي» to phrases that say what the searcher wants. That means «قهوه مختصه», «حبوب قهوة» (coffee beans), or an origin country or bean variety. Add brewing-method and equipment phrases if you sell the gear. Fewer people search them, but those who do are closer to buying.

**Next, the product page.** Near the buy button, show what a buyer checks before paying:

- the roast date, not just the expiry date
- origin and bean variety
- roast level and weight
- ground or whole bean, and which brewing method it suits

Every rise in the share of visitors who buy raises your click ceiling, without touching the bid.

**Then the path to a second order.** Estimate when a bag usually runs out, from its weight and the gap between your customers' orders. A few days before, send a short WhatsApp message with a link to reorder the same bag. If your store supports subscriptions, offer one after the first order. My view is that people subscribe to a coffee they've tasted, not one they haven't.

**After that, bundles.** Two bags from different origins, or a bag with a brewing tool, raise what you keep from the first order. So the ceiling rises before the customer even comes back. See the [product bundles guide](https://mqudah.com/en/insights/product-bundles/).

**Finally, refusals at the door.** With cash on delivery, a buyer can refuse the parcel when the courier arrives. Confirm every order before it leaves the warehouse, and follow the courier closely. A refused order never becomes a returning customer, and it takes the shipping cost with it. See the [cash-on-delivery refusals guide](https://mqudah.com/en/insights/cash-on-delivery-refusals/).

### What not to do

Don't raise your click ceiling on a return rate you haven't measured. If you tell yourself "they'll come back twice" and the sheet doesn't show it, you're paying today for a wish.

And don't chase «كوفي» because its clicks are cheap. A cheap click that doesn't sell costs more than an expensive one that does.

## When is it more than an account tweak?

You've set both ceilings, cleaned up the search terms and added the reminder. The measurement period ends, and a new customer still costs more than they leave. Then the fault usually sits where no single report shows it. It could be the offer, the tracking, or new and returning customers mixed together. Click prices in other sectors are in [Google Ads cost per click in Saudi Arabia 2026](https://mqudah.com/en/insights/google-ads-cpc-saudi-2026/).

That's what I work on in [e-commerce growth](https://mqudah.com/en/services/ecommerce-growth/). I start from the sales sheet, separate new customers from returning ones, and only then judge the ad.

## Questions

### Should I bid on «كوفي» because its clicks are cheap?

A cheap click doesn't mean a cheap order. My reading is that most people who type it want a café nearby, a drinks menu or a name for their own coffee shop. If you keep it, don't buy the bare word. Use phrase match inside buying phrases such as «حبوب كوفي» (coffee beans). Then add negatives such as «قريب» (near), «منيو» (menu) and «وظائف» (jobs).

### How many times does a customer need to come back for the ad to pay?

There's no number that fits every store. Take the customers who first ordered in one month, and follow their delivered orders for six months. Add up what those orders left you, then divide by the number of customers. Build your click ceiling on that figure, not on the one you hope for.

### Should I start with a monthly subscription?

If your store supports it, offer it after the first order, not before. Otherwise a WhatsApp reminder is enough to start with, and it needs no new system.

### Should I advertise on well-known roasters' names?

Not if you don't sell their products. Someone who types a particular roaster's name wants that roaster, and you're paying to talk them out of it. Add those names as negative keywords. Their search volume is still useful in another way, because it tells you where your store stands today.

## Sources

- [Keyword Planner export, Saudi Arabia, Arabic, Oct 2025–Sep 2026](https://ads.google.com/home/tools/keyword-planner/), Google Ads Keyword Planner, pulled by Mohammad Al-Qudah on 7 November 2026
- [About Keyword Planner forecasts](https://support.google.com/google-ads/answer/3022575), Google Ads Help (checked 12 Nov 2026)
- [About negative keywords](https://support.google.com/google-ads/answer/2453972?hl=en), Google Ads Help (checked 24 Oct 2026)
