About Services Cases Blog Tools Reviews Partners Contacts

Grafix case study: Google Ads for a network of metalworking shops from Forbes 250 — leads +17% in six months

Grafix case study: Google Ads for a network of metalworking shops from Forbes 250 — leads +17% in six months

Another case from a tough B2B niche. Grafix is its own network of four large metalworking shops and a separate line selling equipment to other shops: polymerization ovens, paint booths, conveyor systems. Last year the company entered the Forbes 250 ranking, and that adds a certain shine to our cooperation.

By the way, it isn’t the only Forbes 250 Ukraine company I work with right now. Of course, I’m not the reason it got there, but I did lend a hand. With Grafix I work on paid search and SEO.

Contents
  1. About the project
  2. What about the conflict of interest?
  3. In B2B, lead cost is not the main thing
  4. Lead cost grew by 12%. A disaster?
  5. Organic and the overall picture
  6. Breakdown by city
  7. The site: from the red zone to almost green
  8. What’s next: 2027
  9. A recipe for long cooperation
  10. FAQ

About the project

The company operates in four cities: Kyiv, Kharkiv, Lviv and Ivano-Frankivsk. We started with one shop in Kharkiv plus equipment manufacturing.

We have been working together since the start of 2023. Three full years have passed, the fourth is under way. For me that’s already a sign that the advertising is going well. In that time the company ordered external audits from agencies of various kinds. They brought nothing: the audits were about nothing, because those doing them didn’t understand the niche.

And I know this niche well. In parallel, I have also been working for a fourth year with another metalworking shop, where I run both the site and Google Ads.

Google Ads case study: B2B metal processing, 288 UAH ($6.4) per lead over 3.5 years of cooperation Google Ads case study · B2BGoogle Ads case study: B2B metal processing, 288 UAH ($6.4) per lead over 3.5 years of cooperation3.5 years with a B2B metal fabricator: HTML landing page, call tracking, Google Tag Gateway. 757 leads at 288 UAH ($6.4).Sep 13, 2026Read

What about the conflict of interest?

A natural question: aren’t two clients from the same niche a conflict? No.

There are roughly 20–30 competitors in this niche, and I work with two. Neither client knows the other’s data, because I don’t disclose it. And in case studies I don’t publish anything from which you could work out a company’s unit economics: a lot is kept behind a veil of secrecy.

What this gives the clients:

  • Competition with the market. Both clients compete with the other 20–30 players, not with each other.
  • Faster tests. I test different hypotheses on different accounts in parallel. What works is scaled to both projects.
  • Different territories. Grafix has a much wider network and even in Kyiv it covers mostly the right bank, while the other client covers the left. They overlap on a few orders a year.

For the clients it’s no secret. And one more thing: it’s worth looking at competitors, but it’s more important to look at your own business and your own marketing than at somebody else’s greener lawn.

In B2B, lead cost is not the main thing

To an unprepared person who has never worked with B2B and is used to Facebook leads at 5 dollars that don’t turn into deals, the numbers in this case may look expensive. My friends, B2B has different maths.

Lead cost matters: it determines both the budget and the number of leads. But here’s an example, to make it clear, from another project of mine in the same niche.

Out of the blue a request arrives: powder coating of 2,000 m² at 310 UAH (~$6.9) per m² including VAT. That’s 620 thousand UAH (~$13,800) for a single order. Taking into account margin, the ad budget and the fee for my services, it covers all the costs for half a year.

a request for powder coating of 2,000 m², agreed at 310 UAH incl. VAT — client data hidden
a request for powder coating of 2,000 m², agreed at 310 UAH incl. VAT — client data hidden

And that’s the first order. In this niche everything is built on the long-term value of a client. The next order from the same client will repay my services and the marketing for the whole year.

So when order values are measured with five, and sometimes six zeros, a lead for 1,000 or 1,500 UAH (~$22–33) is normal. For powder coating equipment the order value starts from a million UAH (~$22,000).

Lead cost grew by 12%. A disaster?

I compare two half-years: November 2025 — April 2026 and May — October 2026. Leads are calls, forms and email copies according to GA4. Costs are Google Ads, excluding VAT.

If you simply open the ad account and look at the totals, the picture is this: lead cost 919 → 1,031 UAH (~$20 → $23), +12%. Bad, a disaster, what to do, we’re all going to die.

What to do is decomposition. We break the overall figure into small parts and look at each separately.

SegmentBudget, UAHCPC, UAHLeadsCRCPL, UAH
Equipment132,250 → 173,062 (+31%)40.89 → 56.13 (+37%)55 → 89 (+62%)2.17% → 3.60%2,405 → 1,945 (−19%)
Services — search345,872 → 434,338 (+26%)60.35 → 61.58 (+2%)300 → 427 (+42%)6.96% → 7.74%1,153 → 1,017 (−12%)
Services — PMax100,075 → 100,063 (0%)1.51 → 1.50 (0%)274 → 170 (−38%)0.92% → 0.52%365 → 589 (+61%)
Ads in total578,198 → 707,463 (+22%)7.68 → 9.20629 → 686 (+9%)1.71% → 1.68%919 → 1,031 (+12%)

The overall CPC is misleading here: it is pulled down by cheap Performance Max traffic. If you take only search campaigns, the click price is 53.33 → 59.93 UAH (~$1.19 → $1.33), +12%. All amounts in the tables are in UAH; at 45 UAH per dollar, 1,000 UAH is about $22.

Equipment: lead cost −19%

The click got 37% more expensive: from 40.89 to 56.13 UAH (~$0.91 → $1.25). And that isn’t even the highest figure. At the same time we raised the equipment budget by 31%, because it needs to be sold. There were 62% more leads, and lead cost fell by 19% — to 1,945 UAH (~$43).

Specialists who don’t track such things at all, or track clicks, will say that 1,945 UAH per lead is very expensive. Let me remind you: the order value here starts from a million UAH (~$22,000), and conversion to a deal is quite high, because there really aren’t many manufacturers of such equipment.

Services: lead cost −12%

The budget for service search campaigns grew by 26%. The average click price barely changed, but that’s looking globally. Inside there are segments where the click got more expensive and segments that didn’t exist six months ago at all: for example, laser cutting of metal in Kyiv and Lviv. They give a relatively low click price.

The result: leads +42%, site conversion grew from 6.96% to 7.74%, lead cost −12% — 1,017 UAH (~$23). For a lead that potentially brings hundreds of thousands of hryvnias, that’s a good figure.

The conversion rate here is quite high, everything comes down to the expensive click. If you work with Shopping, where a click costs 1–2 UAH (~$0.02–0.04), a click for 60 UAH (~$1.3) can come as a shock. But you should compare things that are comparable.

Performance Max: a dump for this niche

I don’t count Performance Max in the calculations at all. Cheap clicks, cheap “leads”, and few real customers from there. I’m currently reducing the influence of these campaigns to a minimum and, I think, will switch them off completely by the end of the year.

It is PMax that makes the overall lead cost in the account look “worse”: its leads fell by 38%, and the overall average shifted towards the more expensive but quality leads from search.

Where Are the Weak Spots in Your Google Ads Performance Max Campaigns? Google AdsWhere Are the Weak Spots in Your Google Ads Performance Max Campaigns?10 Performance Max weaknesses: tracking, geotargeting, AI features, scaling, traffic cannibalization.Sep 14, 2026Read

Organic and the overall picture

SourceLeadsCR
Ads in total629 → 686 (+9%)1.71% → 1.68%
Organic search193 → 247 (+28%)6.60% → 7.24%
Direct85 → 132 (+55%)1.87% → 1.83%
Total leads907 → 1,065 (+17%)—

Leads from undefined sources (direct) are +55%. There’s no particular point in splitting them between organic and ads here: I do both. So the overall result is this: +17% leads over a comparable half-year.

In such projects it’s harder to pull results out in isolation, with paid search alone: leads are expensive, big budgets are needed. So it’s more logical to connect organic, ads and AI-search visibility and work in a comprehensive way. Marketing has to work as a system. If you only do organic, you lose more than when you work with Google Ads without organic. And together they give a synergy effect: you simply don’t let competitors push you out of the results.

Breakdown by city

SegmentBudget, UAHCPC, UAHLeadsCPL, UAH
Kyiv — search*194,601 → 191,864 (−1%)71.89 → 63.15 (−12%)134 → 149 (+11%)1,452 → 1,288 (−11%)
Kharkiv — search114,718 → 138,720 (+21%)55.29 → 79.82 (+44%)109 → 110 (+1%)1,052 → 1,261 (+20%)
Lviv — search5,718 → 49,310 (×8.6)35.29 → 57.88 (+64%)8 → 72 (×9)715 → 685 (−4%)
Ivano-Frankivsk — search9,464 → 44,541 (×4.7)40.10 → 49.82 (+24%)14 → 60 (×4.3)676 → 742 (+10%)
Brand21,372 → 9,903 (−54%)38.79 → 18.65 (−52%)35 → 36 (+3%)611 → 275 (−55%)

\* Kyiv — together with the old campaigns that the new ones replaced.

The first thing I pay attention to: the click price and its dynamics differ a lot by region. In Kyiv the click got 12% cheaper, in Kharkiv 44% more expensive, in Lviv 64% more expensive. Most likely competition in Lviv is growing. In Kharkiv it may be growing too, or the click simply got more expensive.

An important nuance: in Kyiv I play with manual bids, while in Kharkiv and Lviv the campaigns run on automatic strategies. Hence the different dynamics.

Lead cost by city: Kyiv 1,288 UAH (~$29), Kharkiv 1,261 (~$28), Lviv 685 (~$15), Ivano-Frankivsk 742 (~$16). That’s almost a twofold spread between the cheapest and the most expensive region. Lviv and Ivano-Frankivsk are regions we expanded substantially over six months, and they give the cheapest leads.

Brand. There’s not much to analyze here. The click price fell by 52%, lead cost by 55%. Most likely some competitors stopped advertising on the Grafix brand and let us breathe. Why a brand campaign is worth keeping separate I wrote in the article about the brand campaign.

Key idea: the overall lead cost in the account can grow while each separate segment gets cheaper. Until you break the figure into parts, you don’t know what is really happening.

The site: from the red zone to almost green

Separately, on my own initiative and without involving developers, I rebuilt a site that was barely breathing on WordPress with Elementor.

For now there are only lab measurements. The site was fully in the red zone — there wasn’t a single page that worked efficiently. Now the mobile score on key pages is 90–100, CLS is in the green zone on 19 of 20 pages, LCP is in the yellow zone on most pages and in the green on three.

PageSpeed lab measurements for 20 key pages after optimization
PageSpeed lab measurements for 20 key pages after optimization

Real Chrome data (CrUX) in Search Console updates over 28 days, so for now the old picture is still there: LCP 4.4 s, INP 289 ms, CLS 0.21. When the new data catches up, there will be a separate case study. I’m sure it will show the green zone.

What exactly I did with the site is in a separate technical case study.

Technical Audit and Recovery of the Grafix Website: PageSpeed 22 → 60, Server Response 2.6 s → 0.05 s Технічне SEO · WordPressTechnical Audit and Recovery of the Grafix Website: PageSpeed 22 → 60, Server Response 2.6 s → 0.05 sGrafix’s WordPress site: 141 slow pages, a site archive in the public folder, 15 unnecessary plugins. What I fixed and the numbers I got.Nov 4, 2026Read

What’s next: 2027

The plan for 2027 is fully focused on organic traffic, visibility in AI search and site conversion. The analytical and technical work is already substantial.

The main hypothesis: the expensive click here is partly tied to the site. Click price is formed, among other things, by landing page quality, and until recently the site was extremely inefficient and slow to load. On another project in this niche the click price is much lower, and I think the difference is exactly the site. For now it’s a hypothesis, I haven’t proven it.

If other market factors don’t interfere, in 3–6 months I expect:

  • a lower click price thanks to a fast site;
  • higher conversion of equipment pages — it’s growing, but hasn’t caught up with services yet;
  • as a result, a lower lead cost without extra budget.

A recipe for long cooperation

One of my recipes for long-term cooperation is when nobody gets in your way. Nobody pulls at which keywords we show for. When you are fully trusted and given the work, the client gets the maximum of what I can give. That’s exactly why I rebuilt the site myself, without extra approvals.

When “why did we show for this keyword” or “why don’t I see my ad” starts, such projects are mostly one-off for me. We don’t stay together for long, because I have a different approach to work.

If a business owner has time to review search terms in paid search, other processes in the business suffer. And the company starts to stagnate after that.

Want the same result? Get in touch and we’ll review your project

FAQ on the topic

Frequently asked questions

What is a normal lead cost in B2B Google Ads?
It depends on order value and margin. When a single order brings hundreds of thousands of hryvnias, a lead for 1,000–2,000 UAH pays off with a large margin. It’s worth evaluating deal cost and payback, not lead cost by itself.
What is metric decomposition in advertising?
It’s breaking an overall figure — for example, account-wide lead cost — into components: segments, cities, campaign types. That shows what exactly changed and why, not just the final number.
Does Performance Max work for B2B services?
In narrow B2B niches often not: it gives cheap clicks and leads but few real customers. Search campaigns with clear semantics usually bring higher-quality leads.
Can one specialist work with competitors in the same niche?
Yes, if the clients’ data doesn’t overlap and isn’t disclosed. Useful solutions and proven hypotheses are then scaled to both projects, while they compete with the rest of the market.
Does site speed affect the click price in Google Ads?
Landing page experience is one of the ad quality factors that affect position and price in the auction. A slow and inconvenient site can raise the click price and lower conversion.



Contact

Let's discuss
your project

Leave a request — we'll assess your situation, discuss your goals and find the best solution for your business and budget.

🏆
As a Google Partner, I can help you get a Google Ads bonus of up to 55,000 UAH ($1.2K) for your first campaigns.
❌ I don't work with: gambling · betting · alcohol · tobacco · crypto · grey niches · payday loans
▸ new_project.init()
Discuss a project