HubSpot audit · LCA
This audit follows a client relationship through HubSpot, from first email to repeat work. Marketing performs well. From the moment someone engages, the data that should carry them to a deal, a renewal and a bigger retainer is missing or never entered.
Select a stage to jump to its chapter. Each card shows that stage's health and the main thing that fails there.
Commercial implications
The issues in this audit aren't only about tidy data. They affect revenue LCA can see coming, work people repeat, and money spent on contacts and admin that add nothing. The figures below come from the portal. Where a value is an estimate, it says so.
No build work is needed for these. Each is a few hours in HubSpot settings, or a short task for IT.
Where the record chain breaks
A CRM works when one record links to the next. In LCA's portal almost none of these links exist. That's why marketing can't see revenue and client leads can't see engagement.
Stage 1 · Marketing
356 emails, 213,312 deliveries and steady engagement all year. Marketing is the healthiest part of the portal, apart from list hygiene and one deliverability flag.
A permission-based audience, honest engagement figures, and every event and click recorded as a signal against a known contact and company.
Sends and engagement are strong. The database is bloated with imports, open figures are inflated by bots, and event attendance isn't recorded.
Bars: emails delivered (left axis) · Line: open rate, bots removed (right axis)
*October runs to the 26th. Opens stayed between 19.4% and 23.5% every month.
Open rate by how contacts joined (last 30 days)
80% of contacts came from spreadsheet imports. The largest import (7,239 sends) opens at about half the rate of form sign-ups.
Near-perfect delivery, very low unsubscribes, and clicks well above the norm. The opt-in forms bring in the best audience.
2 spam reports in 213k sends11,445 contacts are billed as marketing contacts. That includes 1,400 hard-bounced, 330 unsubscribed, 89 LCA staff and 196 with no email, plus about 4,900 not emailed in a year.
Likely several thousand removable17% of soft bounces are "sending domain misconfiguration", which points to a gap in thisislca.com's SPF, DKIM or DMARC records. This is a DIY fix for IT.
245 events in 6 monthsMost of the 151 forms are event RSVPs, but no Marketing Events are set up. Who attended a Breakfast Waffle or a conference dinner is never recorded.
0 Marketing Events · 54 campaignsStage 2 · Lead
2,143 contacts clicked an email this year. HubSpot has no way to say which of them are warm, which company they work for, or who at LCA should follow up.
A shared definition of a lead and an MQL, a score that rises with clicks and event attendance, and an owner who gets a task when a client or target warms up.
No lead stages in use, no score, no owner on 88% of contacts, and contacts not linked to companies. Warm signals go nowhere.
Bars to scale against all contacts
86% of engaged contacts have no owner, so their engagement never reaches a person.
11,478 contacts by lifecycle stage
Three in four contacts are labelled Lead by default. Press, Associate and Client are mixed in with funnel stages, so none of them can be reported as a funnel.
No MQL or SQL stage · no lead score9,813 contacts have a company name typed in, but only 9 are linked to a company record. You can't see that five people from one developer clicked the same email.
82 company records in totalKoalify has already flagged them. Merging is a quick win before any scoring or routing goes live.
Koalify installedStage 3 · Deal
21 people created deals this year and £2.71m was won. Many deals enter HubSpot late, skip stages, and carry no contact or company, so the open pipeline doesn't show what's coming.
Deals created at the first conversation, moved through each stage with a current close date and amount, and linked to the client and contacts involved.
One in five wins was created within a week of closing. 95% of open deals are overdue and half have no value. Two people do most of the recording.
Open deals due 27 Oct 2026 – 31 Jan 2027, compared with actual wins
No. The forward pipeline covers a third to two-thirds of what usually closes, and 4 of the 8 deals have no amount. Many wins are added only close to the win.
239 deals won since 2023
Time in each stage doesn't match the overall cycle length. For example, the median time at "Fee proposal sent" is 37 days, which suggests dates were entered after the fact.
Main deal owners, last 12 months. Half of all new deals come from three people.
| Owner | Created | Won / closed | Win rate | Open | Overdue | Pattern |
|---|---|---|---|---|---|---|
| Keziah Harrison | 42 | 22 / 36 | 61% | 17 | 16 | Consistent |
| Sam Emery | 30 | 20 / 35 | 57% | 19 | 17 | Consistent |
| Kirsty Moseley | 21 | 4 / 22 | 18% | 16 | 12 | Active, low conversion |
| Ed Williams | 17 | 5 / 15 | 33% | 4 | 4 | Partial |
| Lucy Donaldson | 12 | 3 / 13 | 23% | 5 | 5 | Partial |
| Helena Carrie | 5 | 2 / 7 | 29% | 15 | 15 | Stale pipeline |
| Jonny Popper | 3 | 1 / 1 | – | 13 (£635k) | 13 | Stale pipeline |
| Nick Bowes | 1 | 0 / 0 | – | 14 (£0) | 14 | No values |
Owners include Design, Insight and Finance staff, so LCA works as a pool of client leads rather than a sales team. The process has to be simple enough for occasional users.
82% of deals have no contact and 85% have no company. Client details sit in free-text fields that are barely filled in.
136 of 153 open deals have no contactLegendre extension is logged twice at £84k each, once won and once lost. Get Living fee extension appears three times, and two of those are still open.
Found in a sample, not a full countAmount, Deal Value, Total Fee Value and Total contract value. Reports will disagree depending on which one is used.
26% of all deals have no amountStage 4 · Renewal
Existing and returning clients brought in 72% of this year's won value. Contracts are extended all the time, but HubSpot holds no end dates, so nobody can see which retainers expire next.
Every won deal records its contract start, end and monthly fee. A renewal deal opens automatically 90 days before the end date, owned by the account lead.
Extensions are logged as one-off deals once agreed. The contract fields exist but are empty, and deals owned by people who've left are never reassigned.
£2.71m won, last 12 months, by deal source
Renewals logged as one-off deals include WOL fee extension (£250k), TCE 2-year extension (£240k), Barratt London extension (£150k) and Mishcon de Reya extension (£110k).
Fields that would drive a renewal process
There's no renewals pipeline or renewal deal type. 18 deals are named "extension" and 10 "retainer", and these were found only by searching deal names.
14 people who've left still own 152 deals, including 16 open deals worth £655k: Francesca Sidoli (£200k), Chris Madel (£166k), Seth Oberlander (£159k) and Freddie Counsell (£125k). Nobody has picked them up.
£655k open with deactivated ownersSince April 2025, HubSpot has logged no meetings, no calls and one email. A new owner would inherit a deal name and an amount, but not the relationship.
0 meetings logged in 18 monthsStage 5 · Upsell
LCA has eight service lines, from consultation to social and creative. Growing a client means knowing which of them they already use. In HubSpot that's missing on nine wins out of ten.
Each client company shows its services, sector, spend and contacts, so account leads can spot gaps and marketing can target, for example, social clients who don't yet use consultation.
Service and sector are missing on about 90% of wins, and few deals link to a company. Upsells exist (retainer increases, social extensions) but can't be measured.
Fields needed for cross-sell and account growth
Only 65 companies have a deal linked, and 17 have more than one. Without these links, revenue per client and services per client can't be reported.
24 won deals with a service recorded, by value
That's £2.27m of the £11.0m won since 2023. The other 215 wins (£10.2m) have no service recorded.
"Berkeley: growth of corp retainer (£6.5k to £10k pm)" and "Increasing retainer by £4k p/m" were tracked as deals, and both were lost. Nothing marks a deal as growth rather than new work, so upsell wins and losses can't be compared.
No upsell deal type71 of the 82 company records have no contacts, so a company page can't show its people, services or history.
87% of companies have no contactsPutting the chain back together
Scoring and forecasting only work once contacts, companies and deals are linked. So the plan starts with clean-up, then rules, then automation.