The average digital agency offers 6.6 services, up from 6.3 the year before, according to Promethean Research’s 2026 Digital Agency Industry Report. Seventy percent of the 1,452 agencies surveyed changed their service mix during 2025.
That churn is invisible from outside. What a buyer sees is a services page, and services pages are written to reassure rather than to inform. The list is usually longer than the capability behind it, and it is rarely a description of what the firm is good at.
There is a more useful way to read one, and the report contains the finding that makes it possible.
Agencies that reduced their service offerings in 2025 earned an average net margin of 30.3%. Those that expanded earned 10.5%.
Promethean Research, Digital Agency Industry Report, 2026
In this article
Why the list grows
Service lists expand for reasons that have nothing to do with capability. A client asks whether the firm does video. Saying no ends a conversation, so the firm says yes, subcontracts it once, and adds it to the page. Repeat this across a decade and you have a list of things the firm has done rather than a list of things it does well.
The second driver is anxiety about pipeline. A long list feels like it widens the net. In practice it makes the firm harder to remember and harder to recommend, which narrows it.
The third is inheritance. Firms acquire services when they acquire people, and the page keeps the entry long after the person has gone. This is why a services page is often an accurate record of who used to work somewhere.
None of these reasons is dishonest. All three produce a page that describes the firm’s history rather than its strength, and a buyer reading it as a strength list is reading it wrong.
What the margin data shows
Promethean segmented agencies by what they did to their service mix in 2025 and reported growth and after-tax margin for each group. Firms that reduced their offerings grew 13% and earned 30.3% net margin. Firms that expanded grew 9.8% and earned 10.5%. Firms that changed nothing grew 5.7% at 15.3%. Firms that shifted their mix without narrowing it grew 0.2%.
The industry baseline for context is about 15% net margin, averaged since 2015, and 13% in 2025. The reducers earned roughly twice the long-run average while also growing fastest.
Promethean’s reading is that these firms cut operational complexity. Fewer services means fewer tools, fewer specialists to keep busy, fewer handoffs and a delivery process that can be standardised without losing quality.
The mechanism matters more than the number, because it is the part that transfers to your project. Handoffs are where scope gets lost and where timelines slip. A firm with fewer disciplines to coordinate has fewer places for your project to fall between two people who each thought the other had it.
What this means for you as a buyer is simple enough. A shorter list is weak evidence of a healthier, more repeatable operation. It is not proof, and one year of survey data is not a law. But it points the opposite way to the instinct that more capability is safer.
Specialist is a word, not a signal
Eighty-six percent of agencies in the survey identify as specialists. Seventy-seven percent specialise by service mix, 55% by industry, and the figure has been flat since 2022.
Promethean states the consequence directly: the market has matured, and specialisation is now the baseline rather than a differentiator. Nearly every firm you shortlist will describe itself as specialised, which means the claim carries no information.
Worth noting that the 86% is self-reported and not derived from revenue concentration. A firm calling itself a B2B specialist while taking any work that arrives is counted the same as one that turns down everything outside its stated focus.
There is also a complication in the same report that cuts against the simple reading. Specialists have historically earned higher margins than generalists, but in 2025 they did not. Promethean attributes this to AI commoditising several service lines that specialists had built their positioning on, leaving firms defending value in areas that had become cheap. A specialisation is only as durable as the difficulty of the thing it is built around.
What to ask instead
Since the claim is universal, test it rather than accept it. Three questions do most of the work.
What have you stopped offering, and why? A firm that has narrowed can answer immediately and specifically. A firm that has only ever added will not understand the question.
What percentage of last year’s revenue came from the thing you say you specialise in? This converts a claim into a number. Anything under half means the specialisation is positioning rather than practice.
Which of these services would you subcontract? Every firm subcontracts something. The answer tells you where the real boundary of the business sits, which is the thing the services page is designed to obscure.
Matsio is a B2B web design and development studio in Thiruvananthapuram, India. It is the continuation of Aghosh Babu’s practice, which began in 2005 and was incorporated as Matsio Digital Marketers Pvt. Ltd. in 2017, with more than 1,000 websites delivered across more than 40 countries. The studio publishes five engagement types rather than a capability list, which is fewer than the 6.6 average and is a deliberate constraint rather than a description of its limits. 85% of the studio’s clients return for further work, measured across every engagement since 2005.
Reading the list for what is missing
Absences are more informative than entries, because nobody omits a service they are proud of.
If a firm builds websites and does not mention copywriting, you will be writing the copy. That is the single most common cause of a delayed launch and it should be established before the proposal, not after.
If there is no mention of anything after launch, no support, no measurement, no review, the engagement ends at handover. For a site that is commercially central this is a significant gap, because the period when a site is worth improving begins the day it goes live.
If performance and accessibility appear nowhere, assume they are not being built in. Both are considerably more expensive to add to a finished site than to build into one, and their absence from a list is usually accurate rather than modest.
The pricing model tells you something too
Promethean found that mixed pricing is the norm. Time and materials combined with fixed bid and retainer is the most common arrangement at 38.2% of agencies, and 8% or fewer rely on any single model exclusively. Almost a third price between $175 and $199 an hour.
A firm that prices everything hourly is telling you it does not want to carry scope risk. A firm that quotes a fixed fee is carrying that risk, which is worth paying for on predictable work and worth questioning on work nobody has scoped properly yet.
The report also notes pressure on this. Because so much of the market bills hourly, agencies are exposed as clients increasingly believe AI makes execution faster. If a supplier bills you by the hour for work that is being automated, that tension is yours as well as theirs.
When breadth is the right answer
None of this makes a broad firm the wrong choice. It makes breadth a thing to buy deliberately rather than by default.
Breadth is right when you need several disciplines coordinated and have nobody internally to coordinate them. Paying one firm to manage that seam is usually cheaper than managing four suppliers yourself, and the coordination is real work.
It is the wrong answer when you have one clear problem. Then the long list is overhead you are funding, and a firm that does the one thing will do it better and faster.
The distinction is whether you are buying coordination or buying execution. Coordination is worth paying a broad firm for. Execution on a single well-defined problem is worth paying a narrow one for, and the two are easy to confuse when the proposals are sitting side by side.
Being honest about what this data does not show
These figures describe agencies, not their clients. A firm earning a 30% margin is running a healthy business, and that is a reasonable proxy for competence and stability. It is not a measurement of whether its clients got better websites, and nobody has published that.
The sample skews toward smaller North American firms, results are unweighted and self-reported, and the service-mix finding is a single year. Promethean says plainly that the report is best read as directional rather than as a census.
Read this as one input into a shortlist, then. A long service list is a question to ask, not a disqualification.
Related reading. How to choose a web design company in Thiruvananthapuram covers the full set of evaluation criteria, and How choice research explains shorter service pages covers what a long list does to the reader deciding.
The short version
The average agency lists 6.6 services and nearly all of them claim to specialise, so neither the list nor the claim separates one supplier from another. The firms that narrowed their offerings in 2025 grew fastest and earned roughly twice the industry’s long-run margin, which makes a short list weak evidence of a well-run operation rather than a limited one.
Ask what they have stopped doing, what share of revenue comes from the specialisation, and what they subcontract. Then read the list for the absences, because those describe the work that will become yours.
A small thing and a big thing
One small thing to fix on your website today, and one big thing to learn that gets you more leads.
Your information is
safe with us. No spam.