Most teams I sit with are not under-spending. They are under-diagnosing. The ads look alive. The reports arrive on Monday. The commercial number — paid orders, closed retainers, cash in the account — does not move in proportion to the money leaving the media platforms.
When that happens, the instinct is to buy more: more budget, more creatives, a new agency, a new tool. That instinct is expensive. If the path from click to cash is leaking, extra spend only buys a louder version of last month. This article is the diagnosis I run before anyone is allowed to scale.
What a leak actually is
A leak is not “the algorithm changed.” A leak is a place where a willing buyer falls out of the journey and nobody can point to it on a page. In practice I see four of them, over and over, in service businesses and e-commerce brands across Nairobi and the region.
The offer is unclear. The ad promises a result the landing page does not name. The buyer has to infer what happens next. Inference is where people leave.
The path is slow. A five-second hero video, a carousel, a chat widget, then a form with nine fields. You would not wait. Neither will they.
The lead definition is vanity. “Conversion” means anyone with an email. Sales never rings half of them. You are scaling noise and calling it demand.
Follow-up is optional. The enquiry sits in a shared inbox until Friday. By then the buyer has asked three competitors. Media buying cannot fix a silent inbox.
Walk the path on a phone
Before you open Ads Manager, click your own ad on a phone that is not on office Wi-Fi. Time the landing page. Count the fields. Try to book the thing you sell. If you hesitate, your buyer will leave. This is not a creative opinion. It is the same usability problem Nielsen Norman Group has documented for two decades: people scan, they do not read, and they punish friction.
NN/g — How People Read on the Web
Then open the conversion settings. If “conversion” is a thank-you page that sales never sees, you do not have a tracking problem. You have a definition problem. Google’s own conversion-tracking documentation is useful once you decide what a conversion is allowed to mean.
Google Ads — About conversion tracking
The one-week diagnosis
Day one: export the last 30 days of spend, clicks, and the events you currently call conversions. Do not decorate the sheet. Three columns are enough.
Day two: sit with whoever closes. Ask which of those “conversions” they would actually ring. Circle the ones that match. That circle is qualified demand. Everything outside it is a leak you have been paying for.
Day three: walk the top three landing paths. Screenshot every step. Write one sentence per step: what the buyer is supposed to do. If you cannot write the sentence, the page cannot either.
Day four: check follow-up. How many enquiries got a human reply in under two hours? If you do not know, that is the leak.
Day five: pick one leak. Only one. Fix the offer line, or cut the form, or define the lead, or install a reply rule. Then wait a week. Do not raise budget yet.
When you are allowed to scale
You scale when qualified conversions move with spend, and close rate holds. Not when ROAS looks pretty for four days. Not when a vendor says the learning phase needs more money. If you want a number to watch while you wait, use the three-number scorecard in the next essay — spend, qualified demand, close rate.
If you are an SME still assembling a site that can take a lead, start there. A leaky path with no website is not a media problem. It is a business that is not ready to buy attention. The SME desk on this site lets you mock the page and the Google ad before you spend. Use it. Then send the form if you want the real thing.
I have watched founders raise budget three times on a path I walked in four minutes on a cheap Android. The landing page took nine seconds. The form asked for a company registration number. Sales was in a WhatsApp group that nobody opened on weekends. That is not a media story. That is an operations story wearing an ads costume.
Write the leak in one sentence. Put it at the top of the Monday note. If you cannot write the sentence, you are not ready to scale. You are ready to diagnose. Also read the scorecard essay so you know what “better” means in numbers, not in feelings.
Also read: A founder scorecard with three numbers