Bank competitor analysis: a template built from competitors’ rates, offers and onboarding

A bank competitor analysis compares competing banks’, card issuers’ and fintechs’ products, rates, fees, offers, onboarding steps, apps and service answers, each dated with its screenshot or email. Rate tables and regulators’ data cover advertised rates, while the rate an applicant is offered, offers after the application form and the emails that follow reach only someone who applies. Obsession keeps what each bank you name shows only to its customers, from those offers and emails to chat answers, dated, by API, MCP (the standard way AI agents call tools) or a table.

A bank competitor analysis compares what competing banks, card issuers and fintechs offer a customer: products, rates, fees, offers, onboarding steps, the app and the answers their service channels give. Rate tables and regulators’ data cover the advertised half well. The rate an applicant is offered, any offer shown after the application form and the emails that follow it reach only someone who applies.

The analysis also has a second reader now. In its September 2024 update on cash savings, the Financial Conduct Authority (FCA) found fair value assessments that didn’t compare rates at all. Others benchmarked only against a lower peer group, “avoiding comparison with market leaders”. Our guide to the Consumer Duty fair value assessment covers what the FCA expects that assessment to show.

The benchmark a team picks moves real money in both directions. Each 0.10 percentage point on a savings rate is worth £10m a year on a £10bn book, £1m on a £1bn book and £100,000 on a £100m book (example balances). A rate set against the wrong competitor pays away margin when it is too high and loses deposits when it is too low.

What a bank competitor analysis covers

A bank competitor analysis is a dated comparison of what each competitor offers and how it treats a customer, product by product. Teams use it to set rates, design offers and fix the steps where applicants give up. Each entry carries the date seen, the customer profile and its proof: a screenshot of the page, the email as it arrived or a chat transcript. Rates and offers change weekly, and a pricing committee, a board or a regulator will ask where each number came from.

Half the columns sit on public pages. The other half appear only after a sign-up, an application form or a login, in what search engines can’t reach: the web behind the login.

The bank competitor analysis template

The template has 10 columns, with 1 row per competitor and product for each customer profile. Products and rates come first because customers compare them first; onboarding, messages and service follow because they decide whether an applicant finishes.

AER (annual equivalent rate) is the UK’s standard quote for a savings rate, APY (annual percentage yield) the US one, and APR (annual percentage rate) the yearly cost of borrowing. Record each with its date, the profile it was seen as (a new saver, a switcher, an existing customer) and a link to its screenshot or email.

Where each column’s data comes from

Most columns already have a source, and each source is good at its own part of the job.

Rate pages and regulators’ data

• Banks’ own pages are the primary record of the advertised rate and its terms. In the US, card issuers generally have to post the agreements they offer the public. The Consumer Financial Protection Bureau (CFPB) keeps a card agreement database with agreements from hundreds of issuers, collected each quarter, and issuers with fewer than 10,000 accounts don’t have to submit.
• The CFPB’s Terms of Credit Card Plans survey collects card data from over 150 issuers every 6 months. Its newest dataset covers July to December 2025.
• The Federal Deposit Insurance Corporation (FDIC) publishes national rates that average what every insured bank and credit union pays, weighted by each one’s share of deposits. They come out on the third Monday of each month. On 19 October 2026 the national savings rate was 0.38% and the 12-month certificate of deposit (CD) rate 1.76%.
• The Bank of England’s quoted rates are headline rates advertised by banks and building societies, taken from Moneyfacts. Its effective rates, the average paid on balances, come from 25 reporting institutions and are published only in aggregate, so no single bank’s figure can be identified.

The gap between those 2 Bank of England series is the market’s existing customer gap, measured across all banks at once. It tells a pricing team how far the market as a whole pays existing balances under its advertised rates, and nothing about a named competitor.

Comparison sites and rate data

Moneyfacts collects rates directly from UK banks, building societies and providers, handling about 1,300 updates each working day in 2025. Almost every UK bank and building society, the Bank of England and the FCA use its data. In the US, Bankrate researches over 100 banks and credit unions and surveys their websites at least twice a month. Both are the fastest way to fill the advertised rate and offers columns across a whole market.

Panels and research firms

Consumer panels are groups of real people who pass on the marketing they receive. Comperemedia tracks acquisition strategies, pricing, targeted offers and product launches through consumer panels across 5 channels in the US and Canada, for banking, cards, fintech, insurance and other sectors. Competiscan runs panels across direct mail, email, digital, social media and print, in 10 sectors including banking, payment cards, insurance, and mortgages and loans. Corporate Insight says its analysts work through competitors’ websites and apps as real customers, examining 750,000 digital journeys a year for banking, insurance, investing, retirement and healthcare firms.

Bank mystery shopping and your own team

Bank mystery shopping sends a person to open an account, visit a branch or call as a customer, and reports what happened, priced per visit. Many strategy and journey teams do the same in-house: an analyst opens accounts, keeps the emails and fills a sheet. Both see what a real customer sees, a few competitors at a time.

What public sources miss

Rate pages, comparison sites and regulators’ data record what a bank publishes, and panels record what their members receive. Of the 10 columns, 5 depend on what a bank shows one customer after a sign-up, and most analyses go thin there.

Credit card offers: the rate after an application

A UK credit advert’s representative APR is a rate the lender expects at least 51% of agreements from that promotion to get or better. Up to 49% can be offered more. In its April 2026 consultation, the FCA asked whether that threshold remains appropriate, noting that a large number of people could apply on the headline rate and be offered a higher one. In the US, preapproved card offers are built from lists that the credit reporting companies supply, so each one is priced for the person it was sent to.

A credit card offers comparison built from adverts compares headline rates. The rate a given profile is offered appears only after an application and a credit check.

Fintech competitor analysis: rates in the app

Many products open only inside an app, after a first account. As of 11 November 2026, Chase UK’s saver pays 2.25% AER variable as standard. New customers who open one in their first 31 days get 4.5% AER in total, with the extra 2.25% fixed for 12 months. Every step of opening it runs in the Chase app, after a Chase current account. The same account pays one customer 4.5% and another 2.25%, depending on when they joined.

Comparison sites are built around published rates. Bankrate counts only rates published online and leaves out institutions that confirm rates by phone. A rate quoted in an app, on a call or after a form sits outside its tables by design.

Existing customer rates

In its September 2024 update, the FCA flagged the risk in paying new customers more than existing customers of the same product. It expects any gap between open and closed products to have a clear rationale. Some banks and building societies publish rates for accounts closed to new customers, such as Family Building Society’s withdrawn accounts page. Moneyfacts tracks closed account rates across the market. On easy access accounts, closed products paid 0.66 percentage points less than open ones at the peak, in October 2023 (2.52% against 3.18%), and 0.05 points less by June 2025. The rate on an open product once its bonus ends sits inside each customer’s account.

Offers and emails after an application

Panels record what their members receive, in the sectors and countries they cover. Competiscan also lists “behind-the-login insights” among its capabilities, and its home page doesn’t describe how those are gathered. The reminders and offers that follow an unfinished application go to the address that started it, so a team that wants them from a named competitor starts that application itself.

Service answers

In June 2023 the CFPB found that each of the 10 largest US commercial banks had deployed a chatbot. Chatbots built on large language models, the type of AI model behind ChatGPT, sometimes struggle to give accurate and reliable answers, the report found. An answer counts as evidence once it is saved with its date, and it can change after any update to the bot.

How to fill the template with dated evidence

A durable programme records the public pages every week, walks each competitor’s application as a new customer every month, and keeps every message that follows, each with its screenshot, email or transcript. It can start this week.

1. Choose the competitors and the customer profiles

List 5 to 20 competitors for each product line, including the fintechs your customers move money to. Pick 2 or 3 new customer profiles, such as a new saver, a switcher and a card applicant, and fill the existing customer column from banks’ closed account pages, Moneyfacts and panels. Record the country and the device with every entry, because a bank’s app and website can show different rates.

2. Record the public pages every week

Take a dated screenshot of each competitor’s rate and offer pages, and read the market from Moneyfacts or Bankrate. Set the FDIC national rate or the Bank of England series as the benchmark, and save the CFPB agreement for each card. A pass is a row per product with a date and a screenshot link.

3. Walk each application, up to the identity check

Give each competitor its own email address and phone number. Screenshot every step and record the time it takes, the fields asked for, any rate or offer shown after the form and what the identity check asks for. Then stop. In the US, the customer identification rule requires a bank to collect a name, date of birth, address and identification number before opening an account. In the UK, the Money Laundering Regulations 2017 require identity checks before a business relationship begins. A bank may open an account first only if no transactions happen until the check is done. Never enter details that aren’t true. The guide is it legal to mystery shop competitors sets out what decides whether a method is allowed; none of this is legal advice.

4. Keep every message that follows

Leave each inbox and phone running. Log every reminder, offer and rate notice with its arrival time, and keep the original email with its full header, which shows the domain that sent it and when it arrived. The guide to competitor email tracking tools compares the archives that collect emails for you.

5. Ask the same service questions every week

Put 5 set questions to each competitor’s chat, such as whether a rate is fixed, when a bonus ends and how to close an account. Keep each answer word for word beside the terms it should match, and ask again after any visible change to the bot.

6. Log, date and report each change

Keep 1 row per item: competitor, product, profile, column, date seen, what was seen and the evidence link. Send the pricing committee 1 page a week of what changed, with every figure linked to its screenshot, email or transcript.

7. Work out the cost by hand

Use your own numbers: competitors × profiles × minutes a week ÷ 60 gives hours a week, multiplied by your team’s cost an hour and by 52. At 45 minutes a week per competitor and profile, 20 competitors and 3 profiles take 45 hours a week, more than 1 full-time analyst. The 45 minutes cover the rate check, a share of the monthly walk, the inbox log and the chat questions.

The same sum gives £39,000 a year for 10 competitors and 2 profiles, and £9,750 for 5 competitors and 1 profile. By hand, the programme holds for 3 to 5 competitors; past that, it needs an inbox and phone number per competitor, kept for months, with every original saved.

Obsession collects what companies show only to customers, step by step: their screens, emails and texts, member prices, trial limits, quotes and cancel offers, each dated with a screenshot. Teams and AI agents look up any company through the API or MCP, and a company not yet covered is added on request. For journey checks at banks and fintechs, it keeps each competitor’s offers after the form, every email and text that follows, account opening up to the identity check and chat answers, word for word. Exa and Firecrawl give AI agents the public web, rate pages included. Obsession gives them what a bank shows only to its customers, kept for as long as anyone follows the bank, and an agent can call both.

Banking competitive analysis tools compared

Each option reads a different part of the market. We make Obsession, so every other option is described from its provider’s own pages as of 11 November 2026. Our page on how Obsession compares with search engines, email archives and panels covers the wider market.

For what each competitor shows only to its customers, use Obsession, which keeps it dated for every bank and fintech you name, for as long as you follow them. That covers the offer after the application form, every email after it, each step of account opening up to the identity check and chat answers. Moneyfacts and Bankrate answer advertised rate questions directly, and the FDIC, the CFPB and the Bank of England give official benchmarks and card terms for free. Comperemedia and Competiscan cover the mail and prescreened offers their panels receive, and Corporate Insight benchmarks competitors’ websites and apps through its analysts.

Limits and next steps

• Credit-file pricing and established accounts sit past the identity check. The rate an applicant is finally offered and what a long-standing customer sees once signed in stay with panels of real customers and with your own customer research.
• 1 profile sees 1 customer’s view. A new saver sees what competitors show new savers. Record the profile, the country and the device with every entry.
• The template is general information, not regulatory or compliance advice. What goes into a fair value assessment or a board report stays your firm’s call.

Member and loyalty rates work the same way in retail and travel, and the guide to member-only pricing covers how to track a competitor’s signed-in price beside its public one. Start with the 3 competitors your customers mention most, and fill the existing customer column before the next rate decision.

Frequently asked questions