Research Distribution and Readership: What the Signal Is Worth

Research readership analytics is the measurement of how institutional clients consume a bank’s published research — which notes they open, which analysts they return to, how long they spend, what they download and forward, at both the contact and the institution level.

On its own it measures attention. Joined to what the client holds, what they trade, and which meetings they take, it becomes the evidence behind a coverage decision and behind the broker vote. This piece is about the distance between those two things, and about a shift now underway that makes the usual way of reading the signal less reliable than it was.

How research distribution actually works

Four stages, each of which produces data that the next one depends on.

Entitlements. Before anything is sent, the firm establishes who is permitted to receive what. Entitlements are driven by the client’s payment arrangement, by jurisdiction, and by the firm’s own rules about which clients see which products. This layer determines the denominator for every readership figure that follows.

Distribution. Notes go out — by email, through third-party platforms, through direct feeds into client systems, and increasingly through API delivery into the client’s own infrastructure. One note typically reaches a client by several of these paths at once.

Engagement capture. Opens, reads, clicks, downloads, forwards, time on page, captured per contact and aggregated to the institution.

Attribution. That engagement record becomes one of the inputs behind the client’s commission allocation, alongside meetings, calls, corporate access and trading. The vote is where readership finally converts into a number.

The chain breaks most often between stages three and four. Engagement is captured accurately and then interpreted as though it were value.

Four ways readership data misleads

An open is not a read. Preview panes, mobile notification previews, security scanners and automated archiving all generate opens. A desk reporting open rates without filtering for these is reporting a mix of human attention and infrastructure.

The named reader is rarely the decision maker. Research reaches shared inboxes, gets forwarded internally, and gets summarised verbally in a morning meeting. The portfolio manager who decides the vote may never open anything. Forward tracking helps and does not close the gap.

Volume and value diverge. The analyst with the highest readership across a client base is frequently not the analyst that client’s vote rewards. Broad readership often signals a widely held name where the client reads everyone. Narrow readership on a differentiated call can be worth considerably more.

Absence is ambiguous. A client reading nothing may have deprioritised the franchise, or may be consuming the same content through a channel the firm is not measuring. That second case is becoming more common, for reasons worth taking seriously.

What is changing: the reader may not be a person

In July 2026, Substantive Research and Aiera surveyed 35 large asset managers on machine-readable research. 77% rated broker research the most valuable machine-readable input available to them, and 69% named licensing as the biggest barrier to using it that way.

Read that alongside what is happening on the sell side. Bloomberg reported in January 2025, using Vali Analytics data, that equity analysts at the 15 biggest banks number about 3,000, down from almost 4,600 a decade earlier, with the remaining analysts often covering two or three times as many companies.

Fewer analysts producing more, for clients who want to feed the output into their own models.

The consequence for measurement is direct. When a client ingests research into an internal system — for summarisation, for signal extraction, for a research assistant their analysts query — the human open disappears as a measure of consumption. A firm whose readership reporting rests on opens will see that client’s engagement decline while their actual consumption rises. The desk then reads a falling number and reduces coverage of a client who is using the research more than before.

This is an argument for measuring delivery and entitlement alongside engagement. Those two know what was consumed even when nobody clicked.

The relationship layer holds up through all of this. Mike Carrodus of Substantive Research, writing in July 2026, put it as research remaining “a relationship business with direct analyst access even more crucial now than ever before.” Machine consumption changes what a note is worth as a file. It raises what the analyst is worth as a person.

Why the evidence matters more than it did

Research payment rules have moved twice in the firm’s favour, and in both cases toward needing a defensible record.

The FCA restored payment optionality in PS24/9 and extended it to fund managers in PS25/4; the EU moved the same way through the Listing Act. Bundled payments are available again. Separately, Substantive Research found in July 2025, surveying 40 managers, that 87% expect at least half of research budgets to be client-funded within two years.

Both directions put weight on the same thing. A client paying from their own P&L scrutinises what they are buying. A client passing the cost to end investors documents why. Either way, the broker making the case needs a record of what was delivered and consumed, reconstructable a year later.

Concentration raises the stakes again. Substantive Research, surveying 50 managers with $20 trillion in combined AUM in April 2026, found the top 10 research providers take 55% of an average global budget, with the top 3 taking 24%. Defending a top-three position is worth more than improving a tenth-place one, and the defence is made with evidence.

What to measure instead of opens

Five measures that survive the problems above:

Entitled reach against actual delivery. How many permitted recipients received each note, by channel. This is the denominator everything else divides by, and it is wrong at more firms than anyone admits.

Depth over breadth, per analyst per client. Repeat consumption of one analyst by one institution says more about where the relationship actually sits than total opens across the franchise.

Consumption by channel, including machine channels. API and feed delivery counted alongside email, so that a client shifting to machine ingestion registers as a change in method rather than a decline in interest.

Readership joined to holdings and position changes. Reading on a name the client owns, or has just moved on, carries a different meaning from reading on a name they watch. This join is what turns attention into intelligence.

Time from a readership signal to a follow-up. A desk measuring this learns whether the data reaches anyone in time to be used. For most firms the honest answer is a surprise.

The last two depend on readership living on the same client record as everything else the firm knows. Readership in a reporting tool of its own produces a monthly deck. Readership on a capital markets CRM record produces a reason to call someone this afternoon.

The layer underneath

All of this rests on the distribution list being right. A list maintained by hand decays continuously — people change firms, mandates move, coverage reassigns — and every readership figure computed against a stale list is wrong in a way that cannot be seen from the report.

Blast ties distribution lists to live CRM data and runs campaigns from Outlook, so the list updates when the client record does, and engagement at the institution and contact level flows back to the same record the sales desk reads from. Consumption Hub is where that record becomes the broker vote evidence.

Frequently asked questions

What is research readership analytics? The measurement of how institutional clients consume published research — opens, reads, downloads, forwards and time spent, captured per contact and aggregated to the institution. Its value depends on being joined to holdings, trading and meeting history rather than read in isolation.

How is research readership used in the broker vote? As one input among several. The vote weighs research consumption alongside analyst access, corporate access, sales service and execution. Readership supplies the documentary record of what the client actually used.

Does readership data prove research value? It evidences consumption. Value is the client’s judgment, expressed through the vote. The practical use of readership is to make that judgment harder to make on anecdote.

What happens to readership metrics when clients use AI to read research? Human opens stop tracking consumption. Firms measuring only opens will see engagement fall at clients whose usage is rising. Counting delivery by channel, including API and feed delivery, keeps the picture accurate.

---

Datazoic is a capital markets CRM and sell-side product suite for front-office teams at investment banks — covering equity research, institutional sales, trading, corporate access and investment banking in a single client record.