Designing the Weekly CI Digest That Actually Gets Read
Your weekly CI digest is probably a newsletter wearing an intelligence costume. Here's how to rebuild it as a decision document.
Why Digests Die
A digest is a decision document, not a reading list.
📉 The open-rate death spiral
Week one, the leadership team reads the digest. Week six, they skim it. Week twelve, it sits unopened. The cause is usually invisible from inside: the digest answers 'what happened' but not 'so what' or 'what do we do.' Executives do not have a reading problem; they have a decision problem. A digest that does not end in an implication or an ask teaches its audience that reading it is optional. Volume is never the issue — relevance is.
🧠 Newsletters inform; intelligence decides
A newsletter is a collection of updates with a headline and a subscribe button. An intelligence product is a structured answer to a question someone will act on. The difference shows in the writing: newsletters describe ('Competitor X announced a new line'), intelligence concludes ('Competitor X's new line targets our mid-range accounts; expect pricing pressure next quarter'). Every entry in a digest must pass the test: can the reader act on this, or is it background? Background stays in the tracker.
⏳ The ten-minute cap is a feature, not a constraint
If a digest takes more than ten minutes to read, it will not be read. The cap forces the three-line discipline: signal, source, confidence, implication — done. A ten-minute cap is also a promise to the audience that the analyst did the filtering, so the reader does not have to. When a week genuinely produces nothing material, the digest says so in one line. Credibility is built by the quiet weeks, and spent by the padded ones.
The Five-Section Structure
Signals / What changed / Why it matters / What we should watch / What we recommend.
📡 Five sections, in a fixed order
Section one, Signals this week: the events, three lines each, with source and confidence attached. Section two, What changed: the delta against last week, not a re-list. Section three, Why it matters: the implication for us, one or two lines. Section four, What we should watch: the two or three indicators that would confirm or kill each signal. Section five, What we recommend: asks with owners and dates. Before: 'Competitor X updated its pricing page.' After: 'Competitor X repriced three mid-range SKUs roughly 8% down (source: pricing capture, confidence: high). Implication: margin pressure on accounts A-C; recommend revalidating our quote posture by Friday.'
🎯 The 'so what' discipline
Every signal ends with an implication — no exceptions. The 'so what' is not a summary of the event; it is the statement of what the event changes for us: a bid, a price, a supply relationship, a risk. If an analyst cannot write the implication, the signal is not ready for the digest and stays in the tracker. This rule does more than improve the document; it improves collection, because analysts start capturing events with the implication question already attached.
🧹 Source hygiene keeps the digest honest
Every signal carries a source and a confidence level, even when the source is internal. 'Heard from the sales team' is a source; it just grades lower than a patent filing. Over weeks, the digest becomes a track record — readers learn which sources have been right, and the confidence labels become calibrated. A digest without sources is gossip in a nice layout; with them, it is the company's most reliable picture of the competitive environment.
Cadence, Tools, and the Escalation Rule
Weekly by default, real-time by exception.
🗓️ A fixed slot beats a burst
Pick a day and time — Tuesday 8am, before the leadership meeting — and defend it like a production deadline. A fixed cadence makes the digest part of the operating rhythm instead of an interruption; the audience learns to expect it, and the analyst learns to produce against a real schedule. Digest quality is a schedule problem more than a writing problem: consistent weekly output, even when thin, outranks occasional spectacular editions, because the audience's habit is the product.
🧰 Tools: a tracker, a template, and a short list
The digest itself is a template — the hardest part of writing it is the formatting, so remove formatting from the job. Behind it sits a signal tracker where every entry is logged through the week with source, date, and confidence, so Monday assembly is consolidation, not reconstruction. Distribution is a short list of people who act on it — the digest is a decision document, not a broadcast. When someone asks to be added 'for awareness,' that is the signal the list is too long.
🚨 The escalation rule: three triggers go real-time
If everything is urgent, nothing is. Define three pre-agreed triggers that bypass the weekly digest and fire an immediate alert: a direct competitive attack on a named account, a regulatory or certification event that changes a market's rules, and a leadership or capability move that changes the competitive balance. Everything else waits for Tuesday. The rule preserves the digest's calm authority, and it makes the rare real-time alert land with full force — because the audience knows it only fires for the three things that matter.