The Number That
Doesn't Exist

The UK has three credit reference agencies: Experian, Equifax, and TransUnion. Each one collects data on your financial behaviour, and each one uses a different proprietary model to convert that data into a score. Experian scores you on a scale of 0 to 999. Equifax uses 0 to 1,000. TransUnion — whose data powers Credit Karma — uses 0 to 710. These are not different expressions of the same underlying truth. They are three different opinions, calculated from partially overlapping datasets, using algorithms that weight factors differently.

Your Experian score and your Equifax score can differ by the equivalent of hundreds of points on the same day. This is not a bug. It is the inevitable result of three competing companies building three separate products. Not every lender reports to every bureau. A credit card you opened with one bank may appear on Experian's data but not on Equifax's. A missed payment logged by one agency may not yet have been processed by another. The scores diverge because the underlying data diverges.

What ClearScore, Credit Karma, and Experian's free consumer tools actually are is something the industry rarely states plainly: they are marketing funnels. ClearScore uses Equifax data and shows you a score on a 0–1,000 scale. Credit Karma uses TransUnion data and shows you a score on a 0–710 scale. Experian shows you its own score on a 0–999 scale. In every case, the score is the hook. The product is you — specifically, your creditworthiness profile, which these platforms use to match you with credit cards, loans, and insurance products, earning a commission every time you click through and apply.

The CMA investigated Experian's proposed £275 million acquisition of ClearScore in 2018, concluding that the merger could stifle competition in the consumer credit information market. Experian ultimately abandoned the deal. The CMA's concern was not that consumers were being deceived — it was that the market for showing consumers their credit scores and recommending financial products to them was commercially valuable enough to warrant protecting. That commercial value comes entirely from the affiliate commissions generated when users apply for products.

Your credit file is a structured record of your borrowing history and identity verification data. It is not a comprehensive picture of your financial life. Understanding the difference between what is and isn't included explains a great deal about why the system produces counterintuitive results.

The implication of this asymmetry is striking. You could have £100,000 in savings, a perfect record of paying your rent on time for ten years, and a student loan that you've been repaying diligently — and none of it would appear on your credit file. Meanwhile, someone who has never saved a penny but has held a credit card since they were 18 and always paid the minimum balance would have a far more robust credit history. The system does not reward financial prudence. It rewards borrowing.

The electoral roll is the single most underrated item on this list. Registering to vote is the fastest, easiest, and most impactful thing most people can do to improve their credit file. It is used by lenders as an identity verification tool — confirming that you live where you say you live. People who move frequently, live in shared houses, or have simply never registered are penalised not because they are bad credit risks, but because the system cannot verify who they are.

This is the quiet admission buried in the small print of every credit score platform: the score is built on incomplete data. The lender's affordability assessment — the part that actually determines whether you get the mortgage — requires entirely separate information that the credit bureaux don't hold. Income, employment status, monthly outgoings, childcare costs, existing debt repayments. All of this has to be provided directly to the lender, separately, and assessed against their own criteria.

A hard search is a visible record of a credit application. When you apply for a mortgage, a credit card, or a personal loan, the lender performs a hard search on your file, and that search remains visible to other lenders for 12 months. The conventional wisdom — repeated endlessly by credit score apps — is that hard searches "lower your score." This is technically true but fundamentally misleading about what actually matters.

The reason multiple hard searches concern lenders is not that each search mechanically reduces a number. It is that multiple applications in a short period signal something about your circumstances. Either you are desperate for credit — which suggests financial stress — or you have been rejected by other lenders — which suggests other lenders have already assessed you and found a problem. Both interpretations are bad. The search is a proxy for the story it tells.

The real danger is the application death spiral. Someone gets rejected for a mortgage. They don't understand why — the lender's decline letter is typically vague. They panic and apply to three more lenders in the same week, each application adding a hard search to their file. Each subsequent lender now sees a file with four recent hard searches, which makes the application look worse than it did before the first rejection. The system punishes the exact behaviour that rejection triggers in anxious applicants.

The distinction between hard and soft searches is important and underexplained. A soft search — the kind performed when you check your own credit file, or when a lender does an initial eligibility check — is invisible to other lenders. It has no impact on how your file appears to anyone else. Many lenders now offer soft search eligibility checkers that give you a high-confidence indication of whether you'd be approved before you formally apply. Using these before making any formal application is one of the most practically useful things you can do.

Open Banking — introduced in the UK in 2018 following a CMA order — allows lenders to access your actual bank transaction data with your explicit consent. Instead of relying on the credit bureau's historical record of your borrowing behaviour, a lender using Open Banking can see your real income, your real spending patterns, your rent payments, your savings behaviour, and how your balance moves through the month. It is, in principle, a far more accurate picture of your financial life than a credit file built on borrowing history.

By December 2025, Open Banking had reached 16.5 million active user connections in the UK — a 36% increase over the previous year. For people with thin credit files — the approximately 5 million Britons described by Experian as "credit invisible," including recent immigrants, young adults who have never borrowed, and people who have historically paid for everything in cash — this is potentially transformative. Open Banking allows lenders to assess you on what you actually do with money, not just whether you've borrowed before.

The adoption of Open Banking in lending decisions has been slower than its adoption in payments and account aggregation. Incumbent lenders have calibrated their models on decades of credit bureau data. Switching to Open Banking-based assessment requires rebuilding those models, retraining risk teams, and accepting a period of uncertainty about default rates. The existing system works well enough for the borrowers who already fit the model — which is to say, the borrowers who have already borrowed.

Services like CreditLadder — which uses Open Banking to read your rent payments from your bank account and report them to the credit reference agencies — represent a pragmatic bridge between the old system and the new. They do not require lenders to change their models. They simply add a new category of positive data to the existing credit file. For renters who have been paying £1,200 a month reliably for five years, this is a meaningful improvement to a file that previously showed nothing.

The longer-term trajectory is toward a system where your actual financial behaviour — not just your borrowing history — is the primary input into lending decisions. Whether the incumbents will allow that transition to happen at speed, or whether it will require regulatory intervention, remains an open question. The technology exists. The incentives to deploy it are unevenly distributed.