Your Best Work Is Under NDA: How B2B Firms Prove It Anyway (2026)

Most B2B services and manufacturing firms cannot name their best clients, because the contracts, the clients’ own policies or plain discretion forbid it. The fix is not to make bigger claims to fill the gap. It is to prove the work in other forms: anonymised descriptors, outcome-first case studies, exact numbers, references and people who can explain a project without naming who it was for.

This comes from what we heard in client conversations at Everything Design, across more than 300 calls with B2B founders, marketing heads and leadership teams between July and November 2026. About 45 of those calls ran into the same wall: the proof exists, and it cannot be shown.

Why can B2B firms not name their clients?

B2B firms cannot name their clients because the client usually decides, and most clients say no. A contract manufacturer making products sold under someone else’s brand is bound not to reveal who it makes them for. A cybersecurity firm cannot say which bank it protects. An IT services firm working inside a client’s operations is often treated as part of that client’s own team.

Even willing clients rarely go on camera. One IT services firm told us that when it had asked years earlier, only 3 of its 20 clients agreed to an on-camera testimonial. That rate is a useful planning assumption: if you need named proof, ask far more clients than you need.

How do you describe a client you cannot name?

Describe the client by what makes it impressive, not by its name. A contract manufacturer we spoke to cannot name its customers, so it uses descriptors such as “a national pharmacy network” alongside aggregate claims about volumes and categories, and keeps the actual logos for decks shown in closed rooms.

A good descriptor carries 3 pieces of information: the type of organisation, its scale and what was hard about the work. “A national pharmacy network” tells a buyer more than a logo they might not recognise. Stop before the descriptor becomes identifiable. If sector, region and date together point to one company, generalise one of them.

What does a case study look like without the client’s name?

A case study without the client’s name leads with the outcome the work enabled, not the volume of work delivered. “What we enabled” is a stronger story than “600 people placed”, because the buyer cares what changed for the client, not how busy the supplier was.

Structure it as the problem the client faced, the constraint that made it hard, what was done and what changed afterwards. Our guide to case studies on a B2B website covers the format in full. Two rules matter more when the name is missing.

First, keep raw impact numbers unrounded. An exact figure reads as measured. A rounded one reads as a guess, and an anonymous case study cannot afford to look guessed.

Second, publish one industry-specific case study per vertical you sell into, and drop the ones from unrelated sectors. A buyer in pharma reads a pharma case and skims past a retail one, however good it is.

Can video prove work you cannot name?

Yes. Video lets the people who did the work explain a project without naming the client. A consultant describing the problem, the decisions and the result on camera carries credibility that text cannot, because the viewer can judge the person’s grasp of the detail.

This works for firms whose clients will not appear. The expert becomes the proof. Keep each piece short and about one project, and let the speaker name the sector and the constraint rather than the company. For formats, see our examples of B2B video testimonials.

Which proof points work when names are off the table?

Proof points that are binary and checkable work best: zero failed implementations, or zero data lost in a migration. A buyer can ask about a claim like that in a reference call, which is why it is believed.

Retention is useful but needs explaining. High retention is partly inertia, because switching suppliers is painful. So say why clients stay, not just the rate. A sentence on what clients say keeps them is worth more than the percentage.

Specific capability claims need evidence behind them. Calling yourself innovative needs an actual R&D budget, a team and output you can point to. Without those, the word costs credibility instead of adding it.

Do logos and analyst badges still matter?

Logos matter, but fewer and stronger ones work better than a long wall. Merge customer and investor logos into a short, recognisable set, and show investor logos and customer logos apart, because they prove different things. One says people backed you. The other says people bought from you.

Customer logos often beat analyst badges. In our calls, mid-market buyers trusted customer logos more than analyst recognition. One IT leader ranked credibility signals for us in order: references first, analyst validation second, company values last. Our post on trust signals on a B2B website sets out how to order them on a page.

What if you have almost no proof yet?

When you have almost no proof yet, lean on the team and hold back the pages that need names. A new fund we spoke to had no closed deals. It put the team’s prior experience at the front and kept its advisory page offline until it had 3 or 4 names to show. An empty page is worse than a missing one.

Buyers also look for proof at their own scale. One prospect asked us directly for work done for companies of its revenue size. If your best examples come from much larger or much smaller clients, say what transfers and what does not.

Where does unnamed proof belong on the website?

Unnamed proof belongs next to the claim it supports, not in a separate section nobody visits. Put the anonymised descriptor beside the service it proves, the outcome number beside the capability and the expert video on the page for that sector.

Keep a separate route for the proof you can only share privately. Offer references, closed-room case studies and named logos in the sales conversation, and say on the site that they are available. That tells the buyer the proof exists and that you respect the confidentiality they will also expect from you. For how proof supports a long sales cycle, see how enterprise buyers decide to trust.

Proof is read by more than buyers. See the website visitors who never buy from you.

FAQs

Can I publish a case study without the client’s permission if I remove the name?
Check the contract first. Many confidentiality clauses cover the work itself, not just the name. If the descriptor plus the details would let a reader identify the client, it is not anonymous.

How many clients will agree to a video testimonial?
Fewer than you expect. One IT services firm found only 3 of 20 agreed. Ask widely and offer formats that do not need the client on camera.

Are analyst badges worth showing?
Show them if you have them, but below customer proof. In our calls, references and customer logos carried more weight with buyers than analyst recognition.

Should I round impact numbers?
No. Unrounded numbers read as measured. Rounded ones read as estimates.

If your best work is under NDA and your website shows none of it, book a call.