Increase the signal, reduce the noise = £
Beacon turns an estate's digital noise into signals leaders can act on. What live work across four pubs could become across 240, from review health and listing drift to brand risk and campaign reach.
Beacon runs today on four pubs. This report sets out what it could become across the estate’s 240 pubs, about 180 of them tenanted: the areas of the business it reaches, the problems it takes away in each, and the signal it produces.
The weight of it sits where the estate does, in the tenanted houses. Nothing here is invented: every line is an extension of machinery already running.
What Beacon does today, live on four pubs
- Reads how each pub surfaces in AI answers: the questions we lead, the ones we appear in, the ones we are missing from, with the winning rival named on each.
- Judges each pub’s Google listing against the rivals that beat it, scored on the five parts of a complete listing, refreshed every six hours.
- Holds each pub’s Google and TripAdvisor reviews in one tank, refreshed on a standing schedule.
- Carries a task board where every task names its steps, its owner and a success test a person can check.
- Delivered inside Shepherd Neame’s own Teams. No new logins.
The areas it reaches at estate scale
Widened to the full estate, the same machinery covers six areas of the business. For each: the problems it takes away, and the signal it produces.
The tenanted estate: Greg
We already run the 180 tenanted pubs’ websites on their behalf, and their listing facts flow centrally through Yext. Central digital run for a tenant is not a new idea here: the websites prove it works. What nobody runs is the rest of their digital front: the social pages, the reviews, the Google profiles. That is where the risk and the silence live.
Problems it takes away
- 180 pubs speaking as the brand through licensee-run social pages and profiles, with no central sight and no checks.
- Risky posts, including AI-written ones, going out in a pub’s name and found by luck after they have spread.
- Reviews on tenanted pubs sitting unanswered for weeks, read by every future customer who looks the pub up.
- No early warning that a tenancy is struggling until the trading numbers say so.
Signal it produces
- Every tenanted page and listing read nightly, risk flags on the desk by morning with the post in hand.
- A pub going digitally quiet raised as an early warning. A silent page usually walks ahead of a struggling house.
- Per-tenancy review health, response rate and response time, visible pub by pub.
- Later, the same readings handed to licensees as help: which questions their pub is missing from, what the winning rival does, what to fix first. The same on-behalf-of arrangement the websites already prove.
The managed retail estate: Shane
Problems it takes away
- Campaigns and standards applied house by house, by hand, by busy managers.
- Listing quality varying pub to pub with nobody scoring it.
- No proof of what a campaign actually reached.
Signal it produces
- One campaign published to every managed house in one act, with an approval trail.
- Each pub’s Google listing judged against the rivals that beat it, with the moves to close the gap named.
- Reach figures per pub, so what worked is a number, not a feeling. The ninety-day proof runs here first.
Marketing: Kat
Problems it takes away
- The brand described differently on every channel, pub by pub.
- No sight of how AI describes the pubs when customers ask it where to eat and drink.
Signal it produces
- One voice from one fact spine: a fact changed once and spoken identically on web, listings and Google the same hour.
- Monthly AI-answer readings per pub, with the winning rival named and the reason quoted.
- The top reason each pub loses its AI answers, named from the answers’ own words and cited sources, with the fix to test at the next sweep.
Social media: Ellis
Problems it takes away
- Hundreds of pub pages with no central sight of what is being posted.
- Posting to the estate by hand, pub by pub, page by page.
Signal it produces
- Every post and comment on every page, read nightly, with a drafted response beside anything that needs one.
- A named person pressing approve before anything speaks as the brand.
Web: Lorraine
Problems it takes away
- Website facts and listings drifting apart, caught by complaint.
- The same factual fix done twice or three times, channel by channel.
Signal it produces
- One master copy of every pub fact, feeding web, Yext and Google.
- Drift detected nightly and queued as a fix, not discovered as a surprise.
IT: Paul
Problems it takes away
- Departments buying separate tools, each with its own login and its own data home.
- AI already in use across the estate with no governance at all.
Signal it produces
- One system inside Teams and M365, single sign-on, no new vendor logins.
- Every AI-drafted word approved by a named person before it goes out, logged and auditable.
The board
Problems it takes away
- No single view of how the estate sounds to customers and to AI.
- Listening-and-speaking spend scattered across departments and unmeasured against action.
Signal it produces
- An estate scoreboard: answers we lead, reviews answered inside a day, risks caught, reach delivered.
- One number for what the noise costs, written by the departments themselves in the table below.
What it costs to run
Google and Meta hand their data free to the owner of the page: every review, post, comment and performance number. The keys are access grants, not licences. The watching layer around them runs at about £25 a month at estate scale. The build rides the existing retainer.
What it needs is access: the Google Business Profile application, Meta Business Manager partner access, and a roundup of pub pages that have drifted into private hands. The platforms’ approval queues run weeks, which sets the timetable below.
What we pay today, filled in by us
What each area pays today to listen and to speak, and what got acted on last month. Twenty minutes per line. Whatever it totals is the £ on the cover.
| Area | Who we pay | What for | Cost a year | Acted on last month |
|---|---|---|---|---|
| Tenanted estate: Greg | ||||
| Managed retail: Shane | ||||
| Marketing: Kat | ||||
| Social media: Ellis | ||||
| Web: Lorraine | ||||
| IT: Paul |
Next steps: the close
- This week. I sit with Greg, Shane, Kat, Ellis and Lorraine, twenty minutes each. The table above gets filled in, and their problems go into this report in their own words. Paul’s line fills in when we meet.
- Inside two weeks. The Google and Meta application packs sit ready for signature, so approval day is instant, not the start of paperwork.
- September. This report and the live Beacon go to Paul, then together to Jonathan. It touches six areas of the business, so it needs a sponsor who can see them all at once.
- Approval day. The managed houses go on and a ninety-day proof starts, numbers agreed first: reviews answered inside a day, risky posts caught before they spread, campaign reach in figures.
- Day ninety. The numbers speak and we decide the tenanted rollout, where the weight of the estate sits.
The first step is five twenty-minute conversations. Book me the first one.
What leaders should do next
- Give each area owner twenty minutes to name who they pay, what it costs and what got acted on last month.
- Agree the proof before the work starts: reviews answered inside a day, risky posts caught before they spread and campaign reach in figures.
- Prepare the Google and Meta access packs before approval day, so the platform queues do not delay the proof.
- Name one sponsor who can see the whole estate, not just one department’s piece of it.