Free guide:winning new clients predictably in 2026 · 10 pages, freeGet it now

The team page as a sales page: trust without stock photos

The team page is where an enquiry is decided, and it usually shows a purchased smile. Since 2 August 2026 the picture is a legal question too, and Google's own rating guidelines list invented information about your team as the worst possible verdict. What belongs there instead.

Cover: The team page as a sales page: trust without stock photos

For a service business, the team page is almost always the page opened shortly before the enquiry. Anyone considering working with you wants to know who they will be dealing with. It still gets treated as decoration: four purchased smiles, job titles like "Senior Consultant" beneath them, no responsibilities, no way to reach anyone. Since 2 August 2026 the picture is a legal question on top of that. We will look at three things: what Google's own rating guidelines expect at this point, what the research says about spotting AI faces, and what belongs on a team page that actually carries weight.

Why does almost every prospect end up on the team page?

Because a service is not a product on a shelf. What you sell does not exist before the engagement starts, so the prospect looks for the next best evidence: the people meant to deliver it. That is why this page is rarely the first one visited and often the last one before the enquiry.

The honest caveat comes with it: there is barely any dependable data on this question. The only publicly available survey that asks about the picture directly comes from BrightLocal and is twelve years old. In it, 46 per cent of respondents picked a genuine photo of the owner or a member of staff as the most trustworthy option, 33 per cent picked a generic image and 21 per cent picked no image at all. Where an obvious stock photo was put directly against a real one, 64 per cent chose the real one.

Bar chart on which type of image looks most trustworthy on a service business website: a genuine photo of the owner or a member of staff 46 per cent, a generic image 33 per cent, no image at all 21 per cent. Asked to compare an obvious stock photo with a real one directly, 64 per cent chose the real photo.
Genuine photo of the owner or a member of staff 46 %, generic image 33 %, no image 21 %. Stock photo against real photo in a direct comparison: 64 % for the real one. Source: BrightLocal, "Images Inspire Trust on Local Business Websites", published 21.03.2014, more than 1,600 responses from a US consumer panel. Note: twelve years old, a different market, and it measured a stated view on trust rather than behaviour. We quote it because it is the only survey on exactly this question, not because it is dependable.

A twelve-year-old panel survey is not proof. It is an indication, and above all it is a reminder that you will never answer this question for your own audience with somebody else's number. What you can do instead is further down.

What does Google check on a team page?

This gets more concrete, because Google publishes its rating guidelines. The General Guidelines of 11 September 2025 are the document trained raters use to judge pages. Section 2.5.2 states the expectation plainly: "Every page belongs to a website, and it should be clear: Who (what individuals, company, business, organization, government agency, etc.) is responsible for the website." Section 2.5.3 adds the contact routes, explicitly including the names of the people responsible.

Two passages of the same document are uncomfortable for stock photos and invented profiles. The table of deception types lists one of its own: "The website has deliberately inaccurate or misleading information about the website or content creators to make the website appear trustworthy." Pages like that are to be given the lowest rating. And one step below, in section 5.6: "If you find the information about the website or the content creator to be exaggerated or mildly misleading, the Low rating should be used." Overstating is enough.

Said honestly: these ratings are not a ranking factor. Nobody demotes your page by hand, the verdicts calibrate the systems. The document is still the most precise account you will get of the intent behind judging E-E-A-T, and the checks are simple enough to run yourself. An author profile with a name, a responsibility and one verifiable piece of evidence passes them. Four tiles with first names do not.

Would anyone notice whether the photo is real?

That is the usual counter-argument, and the research answers it uncomfortably. A study in Psychological Science dated 13 November 2023 showed people real and AI-generated faces. The result: the AI faces were judged human more often than the real ones. In Experiment 1 with 124 participants the figure was 65.9 per cent against 51.1 per cent for real faces, and in a reanalysis of earlier data 69.5 against 52.2 per cent. The authors call it "AI hyperrealism". A second finding is worth remembering: the people who made the most mistakes were the most confident they were right.

Two values side by side: AI-generated White faces were judged to be human in 65.9 per cent of cases, real White faces in only 51.1 per cent. A reanalysis of earlier data gave 69.5 against 52.2 per cent. The effect appeared for White faces only.
AI-generated White faces were judged human in 65.9 % of cases, real White faces in 51.1 %. Reanalysis of earlier data: 69.5 % against 52.2 %. Source: Miller, Steward, Witkower, Sutherland, Krumhuber, Dawel: "AI Hyperrealism: Why AI Faces Are Perceived as More Real Than Human Ones", Psychological Science, online 13.11.2023, DOI 10.1177/09567976231207095. Experiment 1 with 124 adults, Experiment 2 with 610 adults. Important: the effect appeared for White faces only, because the models are trained predominantly on them.

The limitation is part of the finding: the effect appeared for White faces only, because the image models were trained predominantly on White faces. So anyone pulling a "diverse team" out of a generator gets not just an invented team, but one that visibly carries the AI bias of its training data.

The practical conclusion is not "it will come out". The practical conclusion is that it will not come out until it does, and that happens where it is expensive. A different person sits at the kick-off than the one on the page. A prospect searches for the name and finds nothing. A competitor takes a screenshot. Betting on staying undetected is the opposite of a brand.

What applies since 2 August 2026?

Since that date the transparency obligations in Article 50 of the AI Act apply. Paragraph 4 reads: "Deployers of an AI system that generates or manipulates image, audio or video content constituting a deep fake, shall disclose that the content has been artificially generated or manipulated." Article 3(60) defines a deep fake as "AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful".

Whether a fully invented portrait falls under this is legally unsettled, because the definition turns on resemblance to existing persons or entities. The clear case is a real colleague placed by AI into a scene that never happened. For the team page the sharper rule is a different one anyway. Under section 5(2) no. 3 of the German Unfair Competition Act, a commercial practice is misleading if it carries deceptive statements about "die Person, Eigenschaften oder Rechte des Unternehmers wie Identität, [...] Befähigung, Status, Zulassung, Mitgliedschaften oder Beziehungen", meaning the identity, qualifications and status of the trader. Austria says the same in section 2(1) no. 6 UWG. A team that looks larger, more experienced or simply different from the real one is not a design decision, it is a statement about what your company can do.

Beneath that lies the floor every website has to meet anyway. In Germany, section 5(1) DDG requires "den Namen und die Anschrift, unter der sie niedergelassen sind, bei juristischen Personen zusätzlich die Rechtsform, den Vertretungsberechtigten", the name, address, legal form and authorised representative. In Austria, section 5(1) ECG asks among other things for "Angaben, auf Grund deren die Nutzer mit ihm rasch und unmittelbar in Verbindung treten können", details allowing users to make contact quickly and directly. A contact form without a single named person is the minimum compliance with a statute, not an offer of trust.

What belongs on a team page that sells?

Not more people, but more that can be checked. Five points make the difference, and none of them needs a photo shoot.

Comparison of a decorative team page and an evidence-based one across five rows. Image: purchased studio shot versus a real photo, a phone will do. Name: first name only versus full name. Role: a title such as Senior Consultant versus responsibility for a specific area. Evidence: adjectives such as experienced and passionate versus a verifiable project or certification. Contact: a general form versus reaching the responsible person directly.
Five rows compared. Image: purchased studio shot against a real photo. Name: first name against full name. Role: title against responsibility. Evidence: adjective against a verifiable project or certification. Contact: general form against direct access to the person responsible. Our editorial judgement, not a measurement. The checks follow sections 2.5.2 and 2.5.3 of the Google General Guidelines of 11.09.2025.

The most important point is responsibility. "Senior Consultant" tells a prospect nothing, "runs the technical side of our automation work" answers the question of why they should write to this person. The second is evidence: a project, a certification, a talk, anything somebody can look up. Adjectives are not evidence, and that is precisely what the rating guidelines mean by "exaggerated". What this looks like at full length we described in Case study pages that sell.

On the picture itself: a phone photo taken in the office beats any studio shot from a database, as long as it shows the person who will actually be at the table later. If you do not want to show photos, or are not allowed to, leave them out and add the evidence instead. A page without pictures is honest. A page with somebody else's pictures is a statement about your company. Where AI is involved, it belongs in the caption, which is what AI disclosure is now for. And because team pages go stale, maintaining them belongs in the same remit as the rest of your content, see content governance.

Three levers for the coming weeks

Walk the page with the rater's question. Is it clear at a glance who is responsible, who covers which area and how to reach that person? If it takes you three minutes to find out, a prospect gives it ten seconds and then leaves.

Replace adjectives with evidence. One line per person that somebody can verify. It is the change with the best ratio of effort to effect, because it needs no photo session and no relaunch.

Measure it against your own audience. How many enquiries mention a person by name? You know that figure from your inbox, it costs nothing, and it tells you more about your team page than any external benchmark.

If you want to know how your team page reads to somebody who does not know you, we are happy to look at it together. Usually it is not the photos that are missing, but two sentences per person. 🙋