Process · working notes
The first 90 days: how I take over a digital function
Condensed from a 29-page digital business review I wrote before starting my current role. Audit first, fix the foundations, then build the things that keep working after the plan is over.
By Lucas Lim Kai Yi, Digital Marketing Executive · · 5 min read
Before I started my current role I wrote a 29-page digital business review of the group I was about to join. Nobody asked for it. I wanted to walk in on day one already knowing where the gaps were, rather than spending the first month finding out.
This is the 30/60/90 plan that came out of it, with the reasoning restored. The client numbers stay in the review, because they belong to the business, not to me. The method is mine and it is the same one I would use anywhere.
The order matters more than any single item. Every phase depends on the one before it, and most of the mistakes I have watched people make in a new digital role come from starting at phase three.
Days 1 to 30: audit, fix, quick wins
The first month is spent finding out what is true. Not what the last person said was set up, and not what the dashboard implies. What is actually firing, on which page, and whether the number at the end of the funnel can be trusted.
In practice that means four audits run in parallel. GA4 and the tag setup, event by event, checked against a real booking or a real checkout. Every Google Business Profile in the group, because a wrong opening hour costs more covers than a bad ad. The reservation funnel and the online-order checkout, walked through as a customer on a phone. And the channel mix: what is being spent where, and what each channel is claiming credit for.
The audit produces a list. The rule for the rest of the month is that anything on the list which is both broken and cheap gets fixed immediately, before any new work starts. Foundations first, because everything in the next 60 days will be measured through them.
Then a handful of quick wins that prove the function can ship. A sticky reservation button on every venue site. Upsell bundles on the order flow. Menu pages structured so search engines can read what is on them and what it costs. None of these are clever. All of them are visible within weeks, and that visibility buys the room to do the slower work.
Days 31 to 60: growth infrastructure
With the foundations trustworthy, the second month builds the things that bring in demand the business does not already have.
The centre of it is SEO content aimed at non-brand demand. A restaurant group already ranks for its own names. The demand it is missing is the customer who has not chosen a venue yet: searching by occasion, by neighbourhood, by constraint. Occasion and location landing pages are built to meet that customer, and my own research on 60 restaurant searches is the reason I prioritise occasion pages over the “best X” fight with food bloggers.
Alongside it, three pieces of plumbing. Email-capture journeys, so the traffic the content earns is not lost on the first visit. Corporate and bulk order funnels, which are usually a large and badly served segment for a group with private rooms and catering. And marketplace optimisation on the delivery platforms, which for many venues is the biggest storefront they have and the one nobody owns.
This phase is infrastructure, not campaigns. It is judged on whether the pages exist, index and convert, not on a launch date.
Days 61 to 90: scale and automation
The third month turns what is working into systems that do not need me in the loop every week.
Loyalty and CRM lifecycle automation come first, because a hospitality business lives on the second visit and most of them leave it to chance. Retargeting is switched on only now, once there is a clean audience to retarget and clean tracking to measure it.
The last piece is the one I care about most: executive dashboards for the three numbers the business actually runs on. Covers per day. Average order value. Brand versus non-brand demand. Deployed, refreshed automatically, and owned by the function rather than assembled by hand each month.
I build these myself, because I have found that a dashboard someone else built for you stops being trusted the first time it disagrees with the finance team. A dashboard whose every number you can trace to its source survives that conversation.
What the plan does not promise
It does not promise a revenue figure at day 90. A plan written before day one cannot honestly do that, and any version of it that did would be guessing. What it promises is that by day 90 the numbers can be trusted, the demand engine exists, and the reporting runs without a person pushing it.
It also assumes the audit findings are acted on. The review was 29 pages because the list was long, and a list that long only shrinks if the business agrees on what comes first. Getting that agreement in the first fortnight is the real day-one job.
The short version
Month one: find out what is true and fix what is broken. Month two: build the pages and funnels that bring in customers who have not chosen you yet. Month three: automate the follow-up and the reporting so it keeps running when the plan ends. Do them in that order.