- Lead-to-sale usually breaks in one of three places: the wrong audience, the wrong offer, or slow and unscripted follow-up.
- Label 50 recent leads with sales and marketing together before blaming either side.
- Without source and stage fields in the CRM, no campaign or salesperson can be judged fairly.
- Fix the dominant breakpoint first, then send CRM outcomes back to Meta and Google.
The CRM is full and the campaign report shows a healthy cost per lead. At the weekly meeting, sales repeats last month’s verdict: the leads are not serious. Few quotations go out, fewer are signed, and nobody can say with confidence which campaign produced the last paying customer.
The usual reaction is to replace the agency or blame the salespeople. Either can be right, but often neither is the whole story. This diagnosis separates the three places where lead-to-sale usually breaks, shows how to tell them apart with evidence you already have, and sets out a one-week audit to run before you change anyone or anything.
The symptoms: what “leads but no customers” looks like
The pattern is rarely one dramatic failure. If several of these signals describe your business, treat it as a structural problem, not a slow month.
- Lead volume is stable or rising, but quotations are not.
- Sales has a word for marketing leads. “Price shoppers”, “time wasters”. Salespeople quietly prioritise referrals instead.
- Many leads are never reached, or do not remember filling in a form.
- The first conversation goes straight to price, then the buyer disappears.
- Follow-up depends on who picks up the lead. One salesperson replies within the hour; another sends a price list two days later.
- Nobody can connect a paying customer to a campaign. The CRM source field is blank or set to “other”.
The last symptom matters most: without it, the argument between marketing and sales stays a matter of opinion.
The three places lead-to-sale breaks
Most cases trace back to one of three breakpoints. Each has a different owner and a different fix, which is why replacing the agency or retraining sales can change nothing.
| Breakpoint | What is really happening | The evidence that points to it |
|---|---|---|
| 1. Wrong audience | The campaign reaches people who will fill in a form but are not buyers: wrong company size, wrong role, wrong location, wrong need | Leads that are unreachable, outside your service area, or asking for something you do not sell |
| 2. Wrong offer | The right people respond, but to a promise that attracts the wrong intent: a discount, a free item, a vague “contact us for more information” | Leads who are real buyers but expect a different product, price level or scope than you provide |
| 3. Slow or unscripted follow-up | Good leads arrive, but the reply is late, generic or inconsistent, and the buyer moves on | Leads that matched your ideal customer, were contacted late or poorly, and went quiet |
Breakpoint 1: the audience is wrong
Ad platforms deliver what they are asked to count. If the only signal a campaign receives is a submitted form, it learns to find more people who submit forms. Pre-filled instant forms make signing up quick, which helps volume but not intent; our diagnosis of why lead form ads bring unreachable leads covers the form settings in detail. Broad targeting, search keywords that also match job seekers, and placements that invite accidental taps show up here too.
Breakpoint 2: the offer attracts the wrong intent
This breakpoint is easy to miss, because the leads look real. The person answers the phone and has a genuine need, just not the one you serve. A campaign headlined “Free consultation” or “Best prices in Riyadh” attracts people looking for something free or cheap. A distributor advertising one entry-level product may attract retail buyers when it wants contractors and resellers. The ad worked; it promised the wrong thing.
Breakpoint 3: the follow-up loses good leads
As a general observation, in many B2B businesses across the Gulf and Egypt the real sale tends to happen after the form: on WhatsApp, on a call, at a site visit or after a formal quotation. That stretch is often unmanaged, with no agreed response time, no standard first reply, no qualification questions and no rule for how many attempts are made before a lead is closed. Good leads can be lost here and recorded in the CRM as “not interested”. Our diagnosis of WhatsApp inquiries that go cold covers this stretch in depth.
Before the audit: make sure you can see the outcome
You cannot diagnose a broken funnel if the CRM does not record what happened to each lead. Two fields need to exist before anything else: source (which campaign, ad or channel produced the lead) and stage (new, contacted, qualified, quotation sent, won, lost, with a lost reason). If they are missing, we suggest adding them now and back-filling the last 30 to 60 days of leads by hand.
Once the fields exist, the outcomes can also be sent back to the ad platforms so they stop learning from form fills alone:
- Meta. Leads from instant forms can be retrieved through the API or webhooks so they reach your CRM in near real time, and CRM data on lead quality can be sent back through the Conversions API (Meta lead ads guide). Meta’s conversion leads guide lists the conditions for optimising toward a later CRM stage: leads from instant forms, at least 200 leads a month, uploads at least once a day, the target stage reached within 28 days of the lead, and a conversion rate to that stage between 1% and 40%. It also recommends storing Meta’s lead ID in the CRM (Meta for Developers).
- Google Ads. Offline conversion import uses the Google Click ID (GCLID) saved with each lead, so a sale closed later can be sent back to the account. For advertisers starting out, Google recommends enhanced conversions for leads, which matches hashed customer data such as email addresses to ad events. Google also states that since 15 June 2026, offline conversion import and enhanced conversions for leads uploads go through the Data Manager API and are blocked in the Google Ads API, so check that any CRM connector or developer integration uses the current route (Google Ads Help; Google Ads Help). Requirements change; check the current documentation before you build.
This is not a precondition for the audit below, but it is usually the first fix that follows it.
The one-week audit
We suggest involving the owner or a senior manager, one person from sales, one from marketing (or the agency), with access to the CRM and ad accounts. Plan for a few hours spread across five working days, and avoid changing campaigns during the week: you are observing, not optimising.
Day 1: pull 50 recent leads and classify them
Export the last 50 leads from paid campaigns, or all of them if you have fewer. For each one, sales and marketing agree together on one of four labels:
- Not a buyer: unreachable, fake, wrong country, job seeker, supplier, or asking for something you do not sell.
- Real buyer, wrong fit: genuine need, but the wrong product, scope, budget level or customer type.
- Good fit, lost: matched your ideal customer, but no quotation or no deal followed.
- Good fit, progressing or won.
Doing this together replaces “the leads are bad” with a count both sides accept.
Day 2: read the pattern
The proportions tell you where to look. A large share of label 1 points to the audience, label 2 to the offer and label 3 to follow-up. Often one label stands out. If the leads cannot be labelled because nobody recorded what happened, that is your finding: the measurement problem comes first.
Hypothetical example (invented figures, not a client). A company labels its last 50 leads: 22 not a buyer, 9 real buyer but wrong fit, 14 good fit but lost, 5 progressing or won.
- What it points to: two problems, not one. Almost half the leads are not buyers, which is an audience problem; but 14 good-fit leads were lost, which is a follow-up problem, and those were people the business actually wanted.
- What to fix first: the follow-up. The 14 lost good-fit leads are the cheapest revenue in the funnel, because the campaign already paid for them; a response standard and a written first reply can be in place within days.
- What not to change yet: the agency or the budget. Tighten the audience on the campaign where most of the 22 cluster, but leave everything else until the next 50 leads have been labelled under the new follow-up, so you can see which change moved the count.
Day 3: check the audience and the offer side by side
For each active campaign, note the optimisation goal, the targeting, the headline promise and the form or landing page. Do the “not a buyer” leads cluster in one campaign, placement or search term? Then read the headline as a stranger would: would your ideal customer recognise themselves, and would a price shopper feel invited? On Meta, check whether the higher-intent form setting is switched on (it is off unless you turn it on, according to Meta’s lead form reference) and whether the form asks even one qualifying question.
Day 4: test the follow-up yourself
Submit a realistic inquiry through your own ad, once during working hours and once in the evening. Record how long the first reply takes, who replies and whether they ask about your need before sending a price. Then check the CRM history of the “good fit, lost” leads: first contact time, number of attempts, and what was sent.
Day 5: trace the last ten customers backwards
Take the last ten paying customers who came from marketing. For each, find the first contact, the campaign, the time to first reply and the time to signature. Compare these with the “good fit, lost” group. The difference between the two groups, in speed, in who handled them or in the offer they responded to, is often the clearest answer the week produces.
What to fix first
Start with the breakpoint that dominates and change one thing at a time so you can see what worked.
- If you could not label the leads: fix measurement first. Make source and stage fields mandatory and agree in one sentence what “qualified” means for your business.
- If “not a buyer” dominates: tighten the audience. Add one or two qualifying questions to the form, switch on the higher-intent form setting where it fits, review search terms and placements, and move the campaign goal away from raw form volume toward qualified leads once the CRM can report them. Expect fewer leads. Our performance marketing work is built around this kind of qualified-lead structure.
- If “real buyer, wrong fit” dominates: rewrite the offer before you touch targeting. State who the product or service is for, the scope you deliver and what makes you different, so the wrong buyer filters themselves out. Remove “free” and “cheapest” framing unless that is genuinely your position.
- If “good fit, lost” dominates: fix the follow-up. Set a response-time standard the team can actually meet, write the first WhatsApp and phone reply (our guide to the first WhatsApp reply and qualification questions has a template), define three or four qualification questions, and agree how many follow-up attempts happen before a lead is closed. This is the core of sales enablement: giving the team the words and the sequence, not just the leads. For Azeidk, that meant fast, professional WhatsApp service replies and an inquiry flow that moves from first message to site visit, alongside formal email replies for partners and suppliers.
Connect CRM outcomes back to Meta and Google at scale once the dominant breakpoint is fixed: better signals from a broken funnel teach the platforms faster but do not repair the funnel.
If spend is rising across channels while revenue stays flat, see the earlier diagnosis why your marketing budget is increasing but sales are not; if the problem starts before the form, with clicks but no conversions, start there.
When an outside view helps
Many companies can run this audit themselves. It is harder when marketing and sales are already in open disagreement and nobody internal is seen as neutral, when the ad accounts and the CRM sit with different suppliers, or when the audit points to the offer or positioning, which is difficult to judge from inside.
Run Day 1 before you book: label your last 50 leads with sales in the room. Bring that sheet and your CRM stages to a business diagnosis session, and we will tell you which breakpoint dominates, whether the agency or the follow-up is the first fix, and which numbers to watch while you change it.
Frequently asked questions
Why are my leads not converting into sales?
Usually because of one dominant breakpoint: campaigns reach people who are not buyers, the offer attracts the wrong intent, or good leads are lost to slow or inconsistent follow-up. Labelling a sample of recent leads with sales and marketing together shows which.
Sales says marketing leads are bad. How do I know who is right?
Export 50 recent leads and classify each jointly as not a buyer, real buyer but wrong fit, good fit but lost, or progressing. The proportions point to targeting, the offer or follow-up, and replace opinion with an agreed count.
How do I qualify leads before they reach sales?
Add one or two qualifying questions to the form, consider Meta's higher-intent form setting (a step where people review and confirm their answers), give sales a written first reply and three or four qualification questions, and record a clear qualified stage in the CRM.
Can I tell Meta and Google which leads became customers?
Yes. Meta accepts CRM lead-quality data through the Conversions API, and its conversion leads goal lists requirements such as instant forms, around 200 leads a month and daily uploads. Google Ads supports offline conversion import and enhanced conversions for leads; since 15 June 2026 Google has routed these API uploads through the Data Manager API. Check current documentation first.
- Meta for Developers, Lead Ads (accessed 2026-10-31).
- Meta for Developers, Page Leadgen Forms (Graph API reference) (accessed 2026-10-31).
- Meta for Developers, Conversions API for CRM Integration (accessed 2026-10-31).
- Meta for Developers, Conversions API (accessed 2026-10-31).
- Google Ads Help, About offline conversion imports (accessed 2026-10-31).
- Google Ads Help, Configure the Google tag for enhanced conversions for leads (accessed 2026-10-31).



