Insights
Research · Positioning & performance

Does sharp positioning pay?

A test of the consulting-positioning thesis on Germany's top 100 digital agencies, scored against four years of financials.

Christoph Kayser ·June 2026 ·4 min read
No.positioning ↔ performance · r = −0.11
not significant · n = 100 · 2022–2025

Across Germany’s top 100 digital agencies, how sharply you’re positioned has no bearing on how you perform — across four years, every size band, and two positioning models. Where a signal shows at all, it runs the wrong way.

Read it right. These are the top 100 — the survivors, ranked by revenue. So the finding is that positioning doesn’t separate the winners. Whether a sharp proposition is what put a firm on the list — or kept others off it — this data can’t see. No low performers are in here.

findings

Three findings

Positioning scored 0–100 from each agency's website and trade-press footprint, blind to its financials, then correlated against BVDW revenue, headcount and growth 2022–2025.

−0.11

No link, full stop.

The headline correlation is indistinguishable from zero (p = 0.28). Sharper-positioned agencies are neither more profitable nor faster-growing.

Pooled · n = 100
−0.01

Flat even on its home turf.

On the 40 comms / creative boutiques — the firm type the thesis was built for — it's flat on the consulting score and on a fairer, agency-native one (r = −0.04). The cleanest null in the study.

Apples-to-apples · n = 40
−0.21

Where a signal shows, it's negative.

Against operating leverage — revenue growing faster than headcount — sharper positioning predicts less of it (p = 0.05; mid-size firms r = −0.35). Sharp agencies scale by adding people.

Tuned lens · p = 0.05
the 2×2

Every agency, mapped

Positioning (horizontal) against financial performance (vertical), split at the medians. If the thesis held, the cloud would tilt up and to the right. It doesn't — it fills all four quadrants. Hover any dot.

102030405060700255075100GOLD STANDARDSCALE PLAYERUNDERVALUED EXPERTSEA OF SAMENESSPlan.Net Germany — pos 24, fin 81Reply - Digital Experience — pos 35, fin 80]init[ — pos 39, fin 75valantic — pos 28, fin 75Exxeta — pos 42, fin 62Valtech — pos 38, fin 58team neusta — pos 43, fin 11Conclusion — pos 31, fin 54intive — pos 36, fin 57SYZYGY — pos 29, fin 61fischerAppelt — pos 31, fin 60Merkle Germany — pos 37, fin 78Appsfactory — pos 36, fin 75GROW Digital Group — pos 30, fin 45pilot Agenturgruppe — pos 27, fin 62dotSource — pos 36, fin 36Interone — pos 33, fin 55MAI Marketing Automation Intelligence — pos 42, fin 24Wiethe Content — pos 18, fin 80Ray Sono — pos 43, fin 62ressourcenmangel — pos 32, fin 46denkwerk — pos 37, fin 46DocCheck agency — pos 53, fin 33AOE Group Holding — pos 34, fin 19achtung! — pos 36, fin 46RYZE Digital — pos 35, fin 50neuland - Büro für Informatik — pos 56, fin 37web-netz — pos 33, fin 49SUNZINET — pos 24, fin 69DigitalNativeAlliance — pos 34, fin 47BRANDUNG — pos 38, fin 28dreifive (Germany) — pos 30, fin 68best it — pos 65, fin 69add2 — pos 31, fin 70ctrl QS — pos 42, fin 35i22 Digitalagentur — pos 32, fin 46port-neo Group — pos 37, fin 76queo — pos 30, fin 48ACTUM Digital — pos 37, fin 91Unic — pos 35, fin 77Forte Digital — pos 22, fin 93la red — pos 32, fin 41format h digital — pos 26, fin 81interactive tools — pos 32, fin 57Mindbox — pos 32, fin 23Vision11 — pos 40, fin 28Ambient Innovation: — pos 46, fin 50Buben & Mädchen — pos 64, fin 70Bitgrip — pos 58, fin 74move elevator — pos 41, fin 77Digital Masters — pos 25, fin 653st kommunikation — pos 23, fin 63ARTUS interactive — pos 31, fin 54avenit — pos 35, fin 43XIMA MEDIA — pos 55, fin 47friendventure.de — pos 50, fin 61Smart Commerce — pos 53, fin 16711media websolutions — pos 28, fin 56.NFQ - Netzfrequenz — pos 36, fin 41JUSTADDSUGAR — pos 48, fin 33SHOPMACHER eCommerce — pos 52, fin 66unitb consulting — pos 37, fin 53von Affenfels — pos 29, fin 51VALID Group — pos 49, fin 58Hatraco — pos 39, fin 56dkd Internet Service — pos 41, fin 27Blackbit digital Commerce — pos 53, fin 70argonauten — pos 36, fin 92Online Marketing Solutions — pos 37, fin 14agnosco.net — pos 29, fin 66dayy — pos 39, fin 99Monday Consulting — pos 46, fin 28TOWA Deutschland — pos 50, fin 35CBE DIGIDEN — pos 16, fin 45360VIER — pos 54, fin 53d-SIRE — pos 22, fin 28Webneo — pos 32, fin 49UEBERBIT — pos 45, fin 48DIXENO — pos 56, fin 15points Gesellschaft für digitale Informationssysteme mbH — pos 37, fin 28The Digitale — pos 51, fin 93Latori — pos 57, fin 3onacy — pos 40, fin 26Wirth & Horn Informationssysteme — pos 61, fin 46KAIPS MARKETING — pos 39, fin 100Lifestyle Webconsulting — pos 45, fin 56njoy online marketing — pos 31, fin 35e-pixler — pos 30, fin 6LightsOn — pos 38, fin 11AnalyticaA — pos 35, fin 4Webmen Internet — pos 21, fin 55THE BRETTINGHAMS — pos 41, fin 32arboro — pos 64, fin 33DREIKON — pos 34, fin 39PANORAMA3000 — pos 48, fin 11DRIVE - Die Medienagentur — pos 30, fin 50Netzbewegung — pos 19, fin 11mobivention — pos 20, fin 16kitconcept — pos 40, fin 70die Medialen — pos 35, fin 10POSITIONING · VISIBLE AUTHORITY SCORE →FINANCIAL PERFORMANCE →
Gold Standard · 24
Sharp positioning and strong financials — the obvious choice.
Scale Player · 27
Generic positioning, strong financials — winning on scale & delivery.
Undervalued Expert · 26
Sharp positioning, lagging financials — not yet monetised.
Sea of Sameness · 23
Generic positioning and weak financials — the commodity middle.

There are two ways to win here: narrow, or wide. Half the strongest agencies are sharp boutiques; half are broad scale players — which is exactly why the line comes out flat.

takeaways

What to take from it

01

The scale path is just as valid.

Scale players win on breadth and size — bigger, broader, billed by headcount (€42M and 312 people on average, vs €26M and 188 for the sharp winners). Two engines, both road-tested.

02

Sharp is a route, not a rule.

A quarter of the field wins with sharp positioning — smaller, focused, strong. But equally sharp agencies sit among the laggards. Sharpness can take you to the top; it doesn't guarantee it.

03

The white space is wide open.

The whole field is undifferentiated (median ~36/100) and unpunished for it. Sharp + productised + leveraged is the unoccupied corner — the breakout as AI commoditises generic build.

method

How it was built

Positioning
0–100, scored by GPT-4.1 across eight dimensions of proposition sharpness, blind to financials, from each agency's website and trade-press footprint. A second, agency-native ruler (platform depth, vertical focus, productisation) agreed — r = 0.61, same verdict.
Performance
From BVDW fee revenue and headcount, 2022–2025: revenue-per-FTE, multi-year CAGR, and operating leverage (revenue CAGR minus headcount CAGR), blended into a percentile composite.
Rigour
Permutation tests and 5-fold cross-validation throughout. Freely re-weighting the model inflated in-sample fit (R ≈ 0.6) but collapsed under cross-validation (≈ 0) — so only simple, pre-specified models are reported.
Caveats. Correlation, not causation, on a single positioning snapshot. Revenue-per-FTE is a margin proxy. Significant results are single borderline p-values among several cuts — directional, not conclusive. And the sample is winners-only: the top 100 are ranked by revenue, so failures never appear. That range restriction biases correlations toward zero — so the claim is "positioning doesn't separate winners among the top 100," not "positioning never matters."

Sources. BVDW Internetagentur-Ranking 2022–2025 · agency websites · Horizont, W&V & Campaign + a ~23,000-article corpus · framework: The Visible Authority.

All insights A dataset to test? hello@christophkayser.de
Christoph Kayser Advisor & Interim Executive
Berlin, Germany
In Em
Insights Positioning & performance
German digital agencies · 2026
Explore Home
Insights
Contact
© 2026 Christoph Kayser · Legal Notice · Information pursuant to § 5 TMG · Christoph Kayser · Berlin · Germany · hello@christophkayser.de · Responsible for content pursuant to § 55 para. 2 RStV: Christoph Kayser. Privacy — this site uses Google Analytics with IP anonymization (Art. 6(1)(f) GDPR). Opt-out.