Google Ads can't save a product nobody is searching for

If too few people search for what you sell, every dollar you put into Google Ads buys very little, however well the campaigns are built.

A few years ago I worked with a company that had built a genuinely interesting product. Novel, well-executed, solving a real problem. Their founder was convinced that paid search would unlock growth.

We looked at the search volume data together. In their entire country, the search terms closest to describing their product were generating around 150 searches per month. Not per day. Per month.

There simply wasn't enough demand in search to build a paid search strategy on. The people who needed this product didn't yet know to look for it - they didn't have the vocabulary for it, and they weren't actively seeking a solution in that category. And a single sale was worth too little to pay for the clicks it would take to find one buyer among so few searches.

No amount of Google Ads spend was going to change that. You can't place an ad in front of a search that doesn't happen.

The company above is an anonymised composite of businesses I have worked with; the details have been changed.

This is a fundamental property of how paid search works - and misunderstanding it is an expensive mistake.

What paid search actually is

Think about how your behaviour differs in two shopping situations.

You need a specific drill bit. You go to a hardware store, walk to the correct aisle, pick up the right size. You knew what you were looking for before you walked in. The store's job was simply to have it and make it findable.

You're spending a Saturday afternoon in a shopping mall. You didn't come for anything specific. Something in a window catches your attention. You go in. You might buy, you might not.

Google Search is the hardware store.

Meta and Instagram are the shopping mall.

Paid search captures people who are already looking for what you sell. The intent exists before the ad. Your ad shows up at the moment someone has decided they want a thing and is figuring out where to get it. Your job is to show up clearly and not waste that moment with a confusing page or irrelevant offer.

Paid social creates the moment of attention for people who weren't looking. You're interrupting - which means the creative, the offer, and the hook have to do the work of creating a need or surfacing a desire the person didn't consciously have before the ad appeared.

Both work. They work differently. Using them interchangeably - or using paid search to do the job of paid social - is where budgets get wasted.

When Google Ads works, and when it doesn't

The distinction comes down to whether demand already exists in search.

Is your business a search business?

Five conditions. For each one, tap the side that sounds like you.

01 The product
02 The demand
03 The market
04 The page
05 The journey

Tap the side that describes your business.

Fig. 1A thinking aid, not a score. Nothing you tap is stored.

When the conditions in the left column are present, paid search is often among the most efficient places to spend, because the buyer arrives with the intent already formed. When you're operating under the conditions in the right column, you're spending to educate people who aren't ready - and the economics rarely work.

A common mistake with channel allocation

The confusion usually comes from treating paid media as a single channel with a single purpose.

"We need more traffic, so we'll run Google Ads."

But the traffic you need from Google Ads - high-intent visitors who are actively looking for your solution - is different from the reach you need from Meta - cold audiences who need to be introduced to the problem you solve.

In practice, many businesses end up doing the opposite of what works:

Running Google Ads for awareness and brand education (usually the weaker use: Search is strongest at capturing intent that already exists. Campaigns on informational queries can work in high-value categories, where one educated buyer is worth a lot, but they need that value to pay for the longer path)
Running Meta retargeting to warm audiences who would likely have converted anyway (partially right, but misses the main opportunity)

The result: Google budget going to informational searches that rarely convert, Meta budget going to audiences that would often have converted regardless of the ad.

That second point has evidence behind it. In large randomised tests of Facebook campaigns, the usual observational ways of measuring ad results often overstated the lift the experiments found, partly because the ads reached people already likely to buy (Gordon et al., 2019). Search shows the same pattern: in eBay's paid search experiments, the ads did little for frequent buyers, who would have bought anyway, and the lift came from new and infrequent users (Blake et al., 2015).

Swap the logic and the economics change completely: Google to capture the people already searching, Meta to reach people who don't yet know they need you.

How to check if your search spend is capturing real intent

Pull your search term report. Don't look at the keywords you're bidding on - look at the actual terms people typed before clicking your ad.

The question to ask for each one: is this the search of someone who has already decided they want this type of product and is comparing options? Or is this someone still figuring out what they need?

The intent signal is usually clear in the phrasing:

The intent is in the phrasing

Low purchase intent terms

what is product
how does product work
is product worth it
alternatives to product
problem solutionswithout specifying the solution type

High purchase intent terms

product buy online
product price
product near me or service city
Competitor brand + product terms
Specific model or specification searches

Fig. 2Search terms as people type them. Dashed boxes describe a kind of search rather than an exact phrase.

High intent terms should get maximum budget allocation. These are the people who are ready.

Low intent terms should usually be excluded or significantly reduced in bid. You're paying to educate people who are likely to click, read, and leave without converting - because they haven't made the decision to buy yet, and a landing page is not where that decision gets made. The exception is a product worth so much per sale that a few of those readers coming back later pays for all the clicks; the next section shows how to check.

Volume is half the test

Search volume on its own doesn't decide whether paid search can pay. The test is expected value: how many searches there are, the share of them you can realistically win, how many of those clicks turn into sales, and how much each sale leaves you after product costs, set against what the clicks cost.

The same 150 searches a month can be worthless or excellent depending on what one sale is worth. In the illustrative example below, both businesses win 20 clicks at $50 each and turn two of them into sales, with a 30% gross margin.

	High-value product	Low-value product
Searches a month	150	150
Clicks won	20	20
Cost of those clicks	$1,000	$1,000
Sales	2 at $20,000	2 at $60
Gross profit on those sales	$12,000	$36
Profit on ad spend (POAS)	12	0.04

Illustrative numbers.

Profit on ad spend (POAS) is gross profit divided by what the advertising cost: above 1, the clicks paid for themselves. The first business should bid for as many of those searches as it can win. The second is buying revenue at a heavy loss, however promising the click-through rate looks.

The question to answer before increasing budget

Before scaling your Google Ads spend, this question needs an answer:

Is the demand already there in search, and is my current spend capturing it efficiently?

If the demand exists and the account is capturing it well - scale. More budget goes to a proven mechanism.

If the demand exists but the account is poorly structured - fix the account first. More budget into a leaky system doesn't help.

If the searches that exist can't pay back their cost at what one sale is worth - the answer isn't more Google Ads budget. It's a different channel or a longer-term approach to building category awareness.

Each scenario calls for a different decision. All three require looking at the data before spending more.

In short

Paid search answers demand that already exists. It has very little power to make more of it.
If nobody is searching for what you sell, no ad budget will fix that.
Volume alone doesn't settle it: a few searches can pay when each sale is worth enough. Check the expected value.
Google Ads = the hardware store (people know what they want). Meta = the shopping mall (discovery).
Using paid search for awareness is a common way budgets get wasted, unless each sale is worth enough to pay for the education.
Before scaling, check: is the demand there, and is my account actually capturing it?

References

Blake, T., Nosko, C., & Tadelis, S. (2015). Consumer heterogeneity and paid search effectiveness: A large-scale field experiment. Econometrica, 83(1), 155–174.
Gordon, B. R., Zettelmeyer, F., Bhargava, N., & Chapsky, D. (2019). A comparison of approaches to advertising measurement: Evidence from big field experiments at Facebook. Marketing Science, 38(2), 193–225.