Two of these professions are represented online by their employers. Three are represented by their own Facebook.
43% of patent attorneys have a page on their firm's website. Of 1,211 certificated enforcement agents, two do — and the eight largest enforcement companies, who employ 630 of them, name none at all.
By Rose Traviss · 28 September 2026
We searched for every one of the 9,169 people on the five registers this site collects — one Google search each, ten results read. One pattern came out of it that none of the registers could have shown, because it is a comparison across professions and every regulator sees only its own.
Some of these professions are presented to the public by their employers. Others are not presented by anybody, so what the public finds is whatever the person put on Facebook.
| Profession | Page on their employer’s site | Social profile in their own name | Ratio |
|---|---|---|---|
| Patent and trade mark attorneys | 42.7% | 4% | 0.1× |
| Costs lawyers | 17.2% | 12% | 0.7× |
| Licensed conveyancers | 2.7% | 24% | 9.0× |
| Notaries public | 0.5% | 8% | 16.8× |
| Certificated enforcement agents | 0.2% | 23% | 139× |
Read the ratio column as: how much more likely is this person to be findable through a personal social account than through their employer. A patent attorney is ten times more likely to have a firm page than a social profile. A certificated enforcement agent is a hundred and thirty-nine times more likely the other way.
That is a thousand-fold spread across five regulated professions doing work that in four cases is a criminal offence to do unauthorised.
The enforcement case
The Certificated Enforcement Agent Register exists for one reason, and HM Courts & Tribunals Service says so on its front page: so that somebody visited by a person claiming to be an enforcement agent can check the claim. That is a consumer-protection function, and it is the only register in this set whose purpose is verification at the door.
Of 1,211 certificates, two of the agents have a page on their employer’s website.
The eight largest employers on the register, between them employing 630 certificated agents:
| Employer | Agents | Agents with a page on the company’s site |
|---|---|---|
| Marston Holdings | 242 | 0 |
| CDER Group | 111 | 0 |
| Bristow & Sutor | 84 | 0 |
| Newlyn plc | 55 | 0 |
| Jacobs Enforcement | 51 | 0 |
| High Court Enforcement Group | 37 | 0 |
| Dukes Bailiffs | 26 | 0 |
| Excel Civil Enforcement | 24 | 0 |
CDER Group, Marston Holdings and Jacobs Enforcement hold the three Approved Enforcement Agency contracts — the Lord Chancellor’s authority to execute warrants issued by the magistrates’ court. Between them they employ 404 of the 1,211 certificated agents on the register, and their websites name none of them.
There is nothing improper in that. No rule requires an enforcement company to publish its agents, and there are obvious reasons — safety among them — why one might choose not to. But it has a consequence worth stating: a householder who does the thing the register was built for, and looks up the name on the card pushed through the door, will not find that person on their employer’s site. They will find a Facebook profile, or nothing.
The conveyancer case is the same and stranger
Licensed conveyancers have the same shape — 2.7% on an employer site, 24% on social — and none of the enforcement industry’s reasons for it. They are office-based property lawyers at firms that market themselves heavily. The CLC register itself is part of the explanation: it populates the employer field from the practice side, so 1,861 of 2,428 conveyancers have no employer recorded at all, and a firm that does not know it employs somebody is unlikely to have written a page about them.
Whatever the cause, the effect is that the most consumer-facing of these professions — the one people meet when buying a house — is the least visible through the firms that employ it.
How these were counted, and two numbers we threw away
Worth setting out, because the first version of this analysis got both columns wrong in the same direction — towards a tidier story.
Counting how often facebook.com appears in somebody’s ten results gives
Marston Holdings’ agents a Facebook presence of 78%. That number is
meaningless. LinkedIn, Facebook and Instagram rank for almost any human name,
and a Facebook result on a search for a common name is usually a different
person with that name.
So every social result here is name-confirmed: the person’s surname and forename must both appear in the result’s URL or title before it counts. That takes Facebook’s apparent hit rate from 1,718 results to 317 — an 18% confirmation rate. Instagram confirms at 34%, X at 58%.
The confirmed figure, 23% of enforcement agents, is the one in the table. The 78% went in the bin. The comparison between professions would have survived either way, because the same test is applied to all of them, but the absolute number would have been four times too high and somebody would have quoted it.
The employer column was wrong for a different reason, and it was worse. A domain counts as a firm’s website if it follows that firm’s people around the search results — if it appears for half or more of them. That test alone gave whitepages.com as the website of eight different firms, archive.org of six, and citma.org.uk — a professional body every trade mark attorney belongs to — of seven. Forty “employer profile pages” for enforcement agents were listings on a people-search site.
A domain now has to pass a second test: it must contain a distinctive word
from the firm’s name, or its initials. irwinmitchell.com for Irwin Mitchell,
jakemp.com for J A Kemp, never whitepages.com for Latimer Lee. That took
the resolved firm websites from 447 to 297 and the enforcement agents’ employer
pages from 40 to 2.
Tightening it then broke it the other way, which is the part worth admitting.
Excluding trade words so that “Chambers IP Ltd” would stop matching
burrowschambers.com also stopped “NMH Costs” matching nmhcosts.co.uk — its
actual website — because once “costs” is excluded there is no word left long
enough to test. Twenty-three real firm sites went that way before a third limb
was added for firms whose name already is an initialism. Every one of the
1,884 employer pages counted here passes a name, initialism or acronym test
against the firm the register names; none rests on the cohort vote alone.
The correction made the finding stronger, which is the uncomfortable part: a wrong number that pointed the same way would have been easy not to notice.
What this does not show
It does not show that these people are hiding, or that anybody has done anything wrong. It measures what a member of the public finds, which is a function of what employers publish, what the professions’ own marketing conventions are, and how common a person’s name is.
It also cannot see a profile that exists and does not rank in ten results. The figures are floors.
And the notary row carries the warning that runs through this whole site: the Faculty Office publishes 772 of its 817 notaries as an initial and a surname, so searches for most notaries are weak by construction and their figures measure the register’s naming convention as much as the profession’s visibility. See methodology.
Method: one Google search per person via Serper, September 2026, ten results
each. Employer pages are counted only where the surname appears in the URL or
title and the domain is one the cohort vote assigned to that firm — a firm’s
people index ranks for “name firm” whether or not that person has a page on it.
Reproducible from serp_analysis.py; tables in data/serp_analysis.md.