AI for insights and analytics
Income can rise on more balances at a thinner margin, and a flat month can hide one product growing and another running off. AI analytics reads the balance, rate and fee files, splits every movement into balance, rate and mix, and answers product finance: where did income come from, and which branches open accounts that never fund.
These are the questions retail banks ask. Each one maps to a measure our tools compute from your files. The AI model chooses the measure and explains the result; the arithmetic is done by our code, and every total is checked.
| The question | The measure behind it | What comes back |
|---|---|---|
| Explain where my income came from this month. | Net interest income bridge | The change split into balance, rate and mix by product and segment, summing to the ledger exactly. |
| Where is margin compressing? | Margin by product | Interest margin by product and segment against the prior period, with the balances behind it. |
| Which segments hold the fees? | Fee income by segment | Fees by product and segment against plan and prior year. |
| Which branches open accounts that never fund? | Opened and never funded | Unfunded openings over openings by branch and channel, with the accounts. |
| Where are deposits quietly leaving? | Balance attrition | Balances leaving against each customer's own baseline, by product and segment. |
| Are we missing plan, and where? | Income against plan | Actual against plan by product, segment and region, with the largest misses. |
Every figure below was computed by a tool from the rows and checked before it was shown. The lines sum; the percentages match; each line opens to the rows that make it.
What is driving the change in net interest income?
Net interest income is up £9.0m on the prior year, split across 3 effects that sum to the change. Balance effect contributes £6.2m, 69% of the movement, while the rate effect is negative on mortgages.
| Balance effect | £6.2m | 69% of the total |
| Mix effect | £1.9m | 21% of the total |
| Rate effect | £900k | 10% of the total |
| The 3 lines sum to | £9.0m | 0 unexplained |
Which products are behind the balance effect?
3 products account for £6.0m of the £6.2m Balance effect, 97% of it. Mortgage balances grew fastest and gave margin away: the rate effect on mortgages alone is minus 2.1 million.
| Mortgages | £3.4m | balances up 9%; margin down 14 basis points |
| Savings accounts | £1.6m | balances up 6%; margin held |
| Current accounts | £1.0m | balances up 3% |
| These 3 are | £6.0m | 97% of Balance effect |
What should product finance take to the pricing committee?
The mortgage margin: balances grew because the rate was cut, and the bridge shows how much of the growth the cut cost. The committee decides the trade; the figures make it explicit.
| Product finance, mortgages | Take the balance and rate split to the pricing committee | This month |
| Regional directors | Review unfunded openings by branch | Weekly |
| Head of product finance | Put the income bridge in the monthly pack | Monthly |
Each measure has one formula, one source and one meaning. They are computed per product and branch and in total, and every one carries an identity that must hold before it is shown.
| Measure | Formula | From | What it tells you |
|---|---|---|---|
| Net interest income bridge | Change = balance effect + rate effect + mix effect, summing exactly | Balances and rates by product and month | Whether income moved on volume, on margin, or on the mix of products |
| Margin by product | Interest income − cost of funds ÷ average balance, by product and segment | Balances; interest; funding rates | Which products grew and gave the margin away |
| Fee income by segment | Fees by product and segment against plan and prior year | Fees file | Which segments hold the fees |
| Opened and never funded | Accounts opened with no funding within 90 days ÷ accounts opened, by branch and channel | Account openings; balances | Which branches open accounts that never become customers |
| Balance attrition | Balances leaving against each customer's own baseline, by product and segment | Balances by customer and month | Where deposits are quietly leaving |
| Products per customer against norm | Products held ÷ customers, against the segment median | Product holdings; customer master | Where the relationship is one product deep |
| Income against plan | Actual income − plan, by product, segment and region | Income; plan file | Where the plan is being missed, and by how much |
| Branch and channel productivity | Funded openings and income per branch and per channel | Openings; income; branch master | Which branches and channels earn |
From your question and the measures declared for retail banking, the model picks the one that answers it, and the period and comparison the question implies.
Deterministic code reads the rows, computes the measure, and checks the identities below. The same question on the same data gives the same answer, every time.
The AI model writes the sentence around the result, naming the product or account behind it. It states no figure that is not in the result, and every figure links to its rows.
| Bridge | Income end − income start − (balance + rate + mix) = 0 |
| Openings | Opened = funded + unfunded + closed within 90 days |
| Balances | Closing balance = opening + inflows − outflows, by product |
The exports retail banks already produce. Column names are mapped once and the mapping is reused. A file is the way in; scheduled delivery and connections to your systems come with the plan, and every source is listed here.
An answer is a list with an owner and a cadence, or it is a chart nobody works.
| Measures | Owner | Cadence |
|---|---|---|
| Income bridge and margin | Product finance | Monthly |
| Unfunded openings; attrition | Regional directors; branch managers | Weekly |
| Fees by segment | Segment heads | Monthly |
The change in interest income between two periods is split into the part explained by balances moving at the old rates, the part explained by rates moving on the old balances, and the part explained by the mix of products shifting. The three sum exactly to the change in the ledger; a residual is shown, never hidden.
Average balances and interest by product, segment and month, funding rates, fees by product and segment, and account openings by branch and channel with the first funding date. Customer identifiers can be replaced with codes before upload on Enterprise.
An account opened with no funding within the window you set is listed by branch and channel, with the opening date. The figure is a fact from the openings and balance files; whether the branch counted it as a sale is a question for the network.
Interest income by product sums to the ledger's total for the period, and the bridge lines sum to the change. Where the product file and the ledger disagree, the difference is reported with the products involved, not absorbed into a line.
Bridge · Deposit flight · Norm penetration · Concentration · Control total · Identity
Asset managers · Commercial banking · Compliance · Construction and building materials · Consulting and advisory · Customer service · Distributors · Education · Finance and FP&A teams · Financial services · FMCG and CPG brands · Foodservice distributors · Freight brokers and 3PLs · Healthcare and med-tech · Hospitality · Industrial distributors · Industrial manufacturers · Insurance brokers · Investment banking · Law firms · Oil and gas services · Pharma · Procurement · SaaS · Sales teams · Shipping and logistics · Sports · Supply chain · Tax and accounting · Telecoms and connectivity · Trading · Wealth managers
Bring a few thousand rows. The data map opens next, every column mapped once, and the first question is answered in minutes. Free, in your browser, no account.