01
Context
What changed on 6 April.
For years, firms have sat on the wrong side of a line. They could send generic financial guidance to everyone, or they could give fully regulated, fully advised recommendations to individuals. Anything in the middle, anything that sounded too specific to one person's situation, risked being treated as advice. The result was an advice gap. Fewer than one in ten UK adults receive regulated financial advice in any given year — a gap the FCA’s new Targeted Support permission was designed to close, after nearly a decade of policy debate.
Under the new permission, an authorised firm can say to a group of customers who share a common situation, "people like you tend to do this," and suggest a specific step. It is neither guidance nor advice — a third thing, narrowly defined, and it is live now. Royal London and Vanguard are through the first-wave approvals. The rest of the market is working out what to do.
02
The shape of the problem
What is actually hard about it.
The framework is elegant on paper. In practice, firms keep running into the same four problems, and they are not the ones the press writes about.
Defining a customer segment that is narrow enough to be useful and broad enough to stay on the right side of advice. Too wide, and the suggestion is meaningless. Too narrow, and it looks like a personal recommendation.
Designing the suggestion itself. What to say, to whom, in which channel, at what moment, without triggering direct marketing rules or inadvertently pushing a firm's own product when something else would suit better.
Proving, after the fact, that the customer was in a better position as a result. Consumer Duty language for an evidence problem most firms already find difficult.
Doing all of the above continuously, at scale, with machine-readable evidence a regulator can inspect on request. Not a one-off report. A live, auditable trail for every suggestion the firm makes.
"A lot of firms are still battling to get their outcomes MI into shape. There's a very real possibility that some firms just won't have the quality of data that's required to prove to the regulator that they can deliver the required standard under Targeted Support."
Jo Cordner · Baringa · FS in Focus
03
Where Abiru fits
A delivery layer, alongside the strategy.
Abiru does not write a firm's Targeted Support strategy. Segment design, operating model and governance decisions belong upstream, with the strategy partners firms already work with.
Abiru is the layer underneath. It takes those designs, runs them at scale, and produces the four specific things the FCA's Variation of Permission asks authorised firms to evidence. Not narrative documents that someone has to write, maintain and re-write each time the product changes, but artefacts the system produces automatically, every time it acts.
For a strategy partner advising firms on Targeted Support, that changes the engagement. Instead of ending at a slide deck and a build plan measured in quarters, the work ends with running software already mapped to the FCA's four evidence requirements, in the client's stack, under their brand. Delivery risk drops, time-to-permission compresses, and the partner's design work lands as operational reality rather than a binder on a shelf.
For a firm running its own Targeted Support directly, the benefit is the same without the deck: the system producing the four artefacts the regulator wants from day one of the permission, continuously from there.
Regulator asks for
Segment definition and exclusion logic
The rules each segment is built from, the cohorts included and excluded, and the record of every change to either. A living document the regulator can query, not a snapshot taken once and forgotten.
Regulator asks for
The Targeted Support journey itself
The specific suggestion, the trigger condition that surfaces it, the channel, the wording, the fallback path if the customer says no. Versioned and re-playable against the exact customer population it was delivered to.
Regulator asks for
Better-position evidence
For each individual suggestion: the signals that triggered it, the reasoning behind it, the alternative path considered, the expected outcome versus not intervening. Machine-generated, per customer, per event, rather than written by a compliance team after the fact.
Regulator asks for
Outcome monitoring and calibration
What actually happened after a suggestion went out. Drift detection when the population changes. Evidence of ongoing calibration rather than a model left untouched. Outcomes MI that a regulator can actually inspect, continuously.
Illustrative sample
One of the four, rendered as the system produces it.
Suggestion
Cash ISA balance £12,400 held 22 months. Surface option to allocate £6,200 to Stocks & Shares ISA at segment default risk profile.
Trigger signals
Cash ratio > 60% of liquid wealth
Tenure in cash ISA > 18 months
No investment products held
Segment: cautious accumulator, 35-55
Counterfactual
No intervention: expected real yield -2.1% over 12 months (AER below inflation forecast).
Expected position delta
+£284 over 12 months, mid-case, post platform fees.
Consumer Duty links
PRIN 2A.5 (price & value) · PRIN 2A.9 (consumer support)