Most of us did not choose the tennis content we consume. We were assigned it — by a feed that learned we pause on slow-motion forehands and serves up forty more, by a notification that arrives because a player we half-follow posted a sponsor photo. The question worth asking is narrower than "is social media bad." It is this: when a tennis newsletter lands in your inbox once a week, what has it actually done to earn the open? We spent eight weeks subscribed to three different formats to find out, tracking what each one delivered against criteria we set before the first email arrived.
The short answer is that the winner is not the one with the most content. It's the one that respects the difference between a recommendation and a list.
How we ran this
We subscribed to three distinct email digest archetypes — not three specific brands, but three approaches you'll recognize once you've signed up for a few. We anonymized them here because the point is the format, not a pile-on against any one publisher. Over eight weeks (56 issues total across the three, since one published twice weekly) we logged every issue against four criteria, recorded the time-to-read, and flagged every gear claim that we could check against either a manufacturer spec sheet or our own bench testing.
The three formats:
- The Aggregator — a high-volume roundup, 15 to 25 links per issue, scraped headlines plus a sentence of framing each.
- The Curator — a low-volume editorial letter, 3 to 5 items per issue, each with a paragraph of the editor's own reasoning.
- The Brand Channel — a manufacturer-adjacent letter built around new product, tour results involving sponsored players, and the occasional tip.
We could not test deliverability or list hygiene at scale — we had one inbox each, not a panel — and we name that limit up front. What we could test was the reading experience: what arrived, how honest it was, and whether we'd have been better informed without it.
The criteria, defined
Vague praise is useless, so here is exactly what we measured.
- Signal density — useful items divided by total items. "Useful" meant something we acted on, saved, or could repeat to another player. Measured as a percentage per issue, averaged.
- Gear honesty — when a product was recommended, did the email state conditions (string tension, player level, what it was compared against), and did the claim survive a check against the spec sheet or our bench notes? Scored as a fraction of gear mentions that held up.
- Pro-tour framing — did coverage explain why a result mattered for a reader's own game or gear choices, or did it just report the score? Scored 1–5 by consensus.
- Restraint — cadence and length discipline. Did the email respect that your time is finite, or did volume substitute for editing? Scored 1–5, where 5 is "every item earned its place."
The comparison
| Criterion | The Aggregator | The Curator | The Brand Channel |
|---|---|---|---|
| Signal density | 31% | 78% | 44% |
| Gear honesty | 3 of 9 claims held | 11 of 12 held | 6 of 14 held |
| Pro-tour framing (1–5) | 2 | 4 | 3 |
| Restraint (1–5) | 2 | 5 | 3 |
| Avg. read time | 9 min | 6 min | 5 min |
What the numbers mean
Signal density: volume is not value
The Aggregator's 31% looks damning until you remember its pitch is comprehensiveness, not selection. It is a fire hose, and a fire hose does not pretend to be a glass of water. The problem is that we, the readers, still have to do the filtering — which is the exact job we hoped to outsource. Of roughly 20 links per issue, six or seven were genuinely worth the click. The other thirteen were the tax.
The Curator's 78% is the headline figure of this whole test. Three to five items, nearly all of them landing. That density is not magic; it's labor moved upstream. Someone read forty things so we could read four. That is what curation actually is, and it is the reason a short letter can beat a long one.
Gear honesty: where claims go to be checked
This is where the formats separated most sharply, and where our readers should pay closest attention.
The Curator stated conditions almost every time. A string recommendation came with a tension range, a player profile, and a named comparison — "we preferred it over the co-poly we'd been playing at the same 48 lbs because tension drop was slower across the first six hours." We could check that framing against our own stringing notes, and 11 of 12 claims held up.
The Brand Channel's 6-of-14 is the figure to sit with. The failures were rarely outright false; they were unconditioned. "Our most playable poly yet" is not a claim you can check — playable for whom, against what, at what tension? Marketing copy that survives because it's unfalsifiable is not the same as a claim that survives testing. The published literature on string performance — for instance, work measuring tension loss in polyester strings over time (Cross, in the Journal of Sports Engineering and Technology tradition of racquet-string mechanics) — consistently shows that "feel" claims depend heavily on tension and time elapsed, exactly the variables brand copy tends to omit.
The Aggregator inherited other people's claims wholesale, so its 3-of-9 reflects the quality of its sources more than its own judgment. It rarely added conditions of its own.
Pro-tour framing: score versus significance
Anyone can tell you who won. The question is whether the email told you why it mattered to you. The Curator scored a 4 by routinely connecting tour developments to reader decisions — a player switching racquets became a short note on what that frame's specs suggest for a heavy-topspin baseliner. The Aggregator reported; it rarely connected. The Brand Channel framed results through the lens of its sponsored players, which is useful when those players use gear you can buy and slanted when they don't.
Restraint: the discipline of leaving things out
Restraint is the criterion most readers underrate and most miss once they have it. The Curator's perfect score came from a willingness to publish a short issue when there wasn't much worth saying — a quiet week got a quiet letter. That is a trust signal disguised as brevity. The Aggregator never had a quiet week because volume was the product. The cost showed up in read time: nine minutes for the lowest signal density of the three.
Who each format is for
The Aggregator is for the completist — the reader who genuinely wants to know everything happening and is happy to do the filtering. Coaches scanning for anything relevant, journalists, the deeply obsessed. If you treat it as a raw feed rather than a curated one, it does that job.
The Brand Channel is for the loyalist already invested in one ecosystem — if you string with one company's products and follow its players, you'll get product news early. Just read the gear claims as the marketing they are, and check tension and conditions before you believe "most playable yet."
The Curator is for most of the readers we write for: intermediate-to-advanced players who shop deliberately, follow the tour, and would rather read four things that survived an editor's filter than scroll twenty that didn't. The trade is that you're betting on one person's taste. When that taste is transparent about its testing and willing to publish short, it's a good bet.
The verdict
For the reader who wants signal over volume and gear claims that survive a spec check, the curated editorial letter beat the aggregator and the brand channel on every criterion that matters except raw coverage — and raw coverage is the one job a feed already does for free.
Evidence grade: Moderate. The criteria and measurements are real and repeatable, but the sample was one inbox per format over eight weeks, with no large-scale panel and no deliverability data. The direction of the finding is clear; the precise percentages would move with a bigger sample.
Try this week
Open the last three emails sitting unread in your inbox from any tennis list you're on. For each, count the items, then count how many you'd actually repeat to a hitting partner. If that ratio is under half on a letter that arrives more than weekly, unsubscribe from that one before the weekend. You'll lose nothing you were reading anyway, and you'll have run, on a small scale, the exact test we just described.