Blog · Lead Generation
How to work back from a won job to the cost per click you can afford
Short answer: systematic lead generation means breaking the path from the ad to the won job into stages and measuring every stage on its own. Click, lead, qualified lead, appointment, proposal, job. Only once you know what share of one stage moves into the next can you work out what a click is allowed to cost. Anyone who simply buys more reach instead widens an existing leak rather than closing it.
The same logic applies in Berlin and in Brandenburg, only with different input figures. In Berlin, high search volume meets a large number of competing bidders; in the surrounding area both are lower. That shifts the cost per click you can afford, not the arithmetic behind it.
All amounts in this article are excluding VAT, so 19 per cent German VAT comes on top. Ranges are industry experience and not a promise. The worked examples are freely chosen, show the logic only and do not describe a real case.
More traffic is the obvious lever, because it is the only one you can move overnight. Yet only half the equation scales: clicks rise immediately, enquiries more slowly, the share of usable enquiries sometimes not at all. Every stage multiplies with the next. If two out of 100 visitors fill in a form and half of those are unusable, double the budget brings twice as many unusable enquiries. Four building blocks have to fit together first.
The weakest building block determines the result. Before you raise budget, work out what a won job is allowed to cost.
The target CPA, meaning the amount a won job may cost in acquisition, is usually estimated. Yet it can be derived, from job value, contribution margin and the share of the contribution margin you put into acquisition. For the third figure, 10 to 30 per cent is a rough guide; the upper end is only defensible with documented repeat business.
A freely chosen example: job value 4,000 euros excluding VAT, margin 35 per cent, so 1,400 euros contribution margin per job. If you put 20 per cent of that into winning new customers, a won job may cost 280 euros excluding VAT. From this figure you work forwards.
| Stage | Transition to the next stage | Allowable cost per unit (excluding VAT) |
|---|---|---|
| Won job | Target figure | 280.00 euros |
| Proposal | 40 per cent turn into a won job | 112.00 euros |
| Appointment | 70 per cent lead to a proposal | 78.40 euros |
| Qualified lead | 50 per cent turn into an appointment | 39.20 euros |
| Lead (form or phone call) | 50 per cent are qualified | 19.60 euros |
| Click on the ad | 8 per cent leave a lead | 1.57 euros |
The cross-check: 1,000 clicks at 1.57 euros cost 1,570 euros and produce 80 leads, 40 qualified leads, 20 appointments, 14 proposals and just under 6 won jobs, so around 280 euros per job.
It gets interesting when you lay real costs per click alongside this. As a rule of thumb from experience, for contested service terms in Berlin they often sit at 3 to 12 euros excluding VAT, and in emergency-callout and consulting niches they run into double figures; your own range is in the Google Keyword Planner and, after two weeks of running, in your own account. Assume a freely chosen 4 euros and the same won job costs around 715 euros, more than twice the allowable amount. That is not a budget problem but a structural one. There are three levers.
If no lever moves far enough, the channel does not fit the offer.
An ad account that optimises for form submissions optimises for form submissions, not for won jobs. This is the most common reason why campaigns look better on paper and worse in sales: the algorithms reliably find people who enjoy filling in forms, the cost per lead falls and the cost per job rises.
Quality needs a definition that two people apply the same way independently of one another. A written list with four criteria has proven itself: the need matches the offer, the budget is realistic, the person has decision-making authority or direct access to the decision, and the timing falls within a sensible window. If a lead meets three of four, it is qualified. That is arbitrary, but unambiguous.
On top of that sits a fixed sequence of statuses: received, contacted, reached, qualified, appointment, proposal, won, lost. Lost gets a reason from a short picklist: price, timing, wrong area, unreachable, competitor. If a third of the losses come down to the wrong area, that is a settings problem in the ad account, not a sales problem.
Per source you need a double-digit number of qualified leads before a comparison shows a tendency. That is not statistically robust, so until then judge at channel level.
Your CRM knows which lead led to a won job; the ad account does not, as long as you do not tell it. This return channel is called offline conversions and consists of four steps.
If you have not captured the click identifier, what remains is enhanced conversions for leads, which you switch on in the conversion settings in Google Ads: they match via the hashed email address instead of via the gclid. You still need to upload the status changes; only the preparatory work in the form falls away, and fewer cases are covered, because the address has to match the account. Expect a delay: there are often weeks between the click and the won job, so the current month is incomplete.
The channel question is usually asked first and should come last. It depends on two figures: active search demand and job value.
| Channel | Suitability | Typical lead quality | Cost per click (excluding VAT, Berlin) | Day-to-day effort |
|---|---|---|---|---|
| Google Search | Urgent need, local services, B2B with search volume | High, the enquiry comes from the person themselves | Rule of thumb 3 to 12 euros, double figures in narrow niches | Medium: search terms, negatives, areas |
| Meta Ads | Offers without search volume, consumer services | Broadly spread, qualification in the form needed | Usually well below Search | High: creatives wear out |
| LinkedIn Ads | B2B with a definable industry, job function and company size | High with precise targeting | Rule of thumb well above Search and Meta, often double figures | Medium |
| Business profile, directories | Local services with a catchment area | Very high, often direct phone calls | No cost per click | Low, maintain reviews |
| Referrals and partners | Almost anything, especially with high job values | The highest of all sources | No cost per click, but time instead | Hard to scale predictably |
| Email sequences | Developing consents you already have | High, the contacts know you | No cost per click, tool costs by list size | Low, the effort is in the build-up |
Hold these ranges against your allowable cost per click: at 1.57 euros a channel with double-digit clicks is out, at 11 euros it comes into play. One sequence that works: first cover the business profile and existing demand through Google Ads, because lead quality is most predictable there. Then create demand through Meta Ads that would not exist without advertising. LinkedIn only once the job value carries double-digit costs per click.
Additional fields usually cost completion rate. Whether they also raise quality depends on whether the field actually qualifies; you test that against each other. Three to five fields have proven themselves: name, email or phone, plus one or two qualifying questions. For lead forms inside the platforms this applies even more strongly: they lower the hurdle, but quality drops, because filling them in costs almost nothing. If you use them, include a qualifying question and measure them separately.
What helps most is not a mandatory question about the budget but a question about the situation: what needs doing, by when, at what scale. A picklist with four to six options is better than a free-text field, because it can be analysed.
Of all the levers, response time has the best ratio of effort to effect, and it is the one most often ignored. A lead who gets a phone call within minutes is in a different state from the same lead on the next working day: they are still at the device and have not in the meantime enquired with two competitors. This is experience from sales processes, not a study.
In practice that means: name a responsible person and a time window, 30 minutes during business hours for example. Deliver leads by email and as a notification to the phone at the same time, not just into an inbox. Plan at least three contact attempts across two channels. And only run campaigns at times when nobody is on duty if a bookable calendar replaces the phone call.
The following section is general guidance on common requirements and expressly not legal advice. For your own case, please bring in a lawyer.
The difference is not mentality but density. In Berlin, high search volume meets a lot of providers, including national ones that buy local terms as well. That drives costs per click up and makes differentiation on the page more important than the bid. In Brandenburg the search volume per town is thin, but the number of competing bidders is smaller, which you can read in the auction insights report in your account: visibility is cheaper, though on its own it rarely fills the chain.
For places such as Potsdam, Oranienburg or Cottbus, separate landing pages are worthwhile, but only if their content actually differs, through service area and journey times for example.
Whether outside support pays off depends on the number of channels and on the job value. The orders of magnitude are in our article on what a performance marketing agency costs, the selection criteria in the checklist for choosing an agency, and we have written up separately how performance marketing in Berlin is set up.
For context: AdSynergy is a performance marketing agency from Berlin, founded and run by Nils Fehr. We work with DTC and ecommerce brands, more than 50 so far, and have managed over 25 million euros in ad budget in total. It works best from around 5,000 euros in monthly ad budget. Our channels: Meta Ads, Google Ads, TikTok Ads, conversion rate optimisation and email through Klaviyo.
The calculation logic above comes from accounts where we steer on contribution margin instead of revenue. For services and in B2B the input figures change, not the logic: instead of basket size and repeat purchase rate, the calculation holds job value and close rate. LinkedIn Ads and the business profile are not part of what we offer; if your case sits there, we say so in the call. We deliberately do not promise results, and no timelines either. What we can commit to is the sequence: make it measurable first, then optimise, then scale.
The entry point is a free, no-obligation 30-minute call. After that you get an audit of your ad accounts with specific recommendations within 24 hours on working days, free and without obligation, even if nothing follows from it. The only requirement is read access to the ad accounts. If you would rather write than talk on the phone, that works through the contact page.
First: work out the allowable cost per won job from back to front, through the transition rates down to the allowable cost per click. This calculation decides which channel is an option at all, and the transition rates are the most effective lever within it. Second: do not optimise for leads but for qualified leads, and for that you need the legally clarified return channel from sales into the ad account. Third: response time is, in our experience rather than on the evidence of studies, one of the strongest levers and it costs no budget. Fourth: Berlin and the surrounding area differ in density, not in logic.
If you want to know where your chain from the ad to the won job breaks: in the free 30-minute call we work out where you stand. After that you get your audit with specific recommendations within 24 hours on working days, even without any later collaboration.
Book a callRead next