← Back to Blog

Behind on Digital? Good. You Get to Skip 15 Years of Bad Decisions.

• Alexandros Fotiadis
digital-transformation smb roadmap strategy automation
Behind on Digital? Good. You Get to Skip 15 Years of Bad Decisions.

If your business has mostly avoided “going digital” until now, there is a real chance this is an advantage, not an embarrassment. We made that argument in the companion post to this one. This post is the practical follow-up: what the first 90 days actually look like, in what order, and for what money.

Three phases. Each phase is short, scoped, and designed to produce value on its own — so that if you stop at the end of any phase, the work you’ve done is already useful.

Phase 1 — Weeks 1–2: audit and quick wins

Goal

Understand what you actually do, in what order, with what tools and what manual effort. Fix the two or three things that are obviously broken.

Work

  • Walk the workflows. Sit with the owner, the office manager, and one front-line person from each function. Map the end-to-end flow of a customer enquiry, an order, an invoice, a supplier payment, a piece of compliance. Write it down in a form anyone can read.
  • Inventory the tools. Every login, every folder, every spreadsheet that touches money, customers, or compliance. Who has access, who doesn’t, what’s abandoned.
  • Quick-win list. Identify two to four things that are obviously broken and obviously cheap to fix: a shared inbox that is actually someone’s personal email, a compliance filing still done by hand that has a reliable automation, a spreadsheet being emailed around that should be a shared document.
  • Decide the target stack at the category level. Email and identity, document storage, CRM, accounting, communications. Not specific products yet — just the categories you will consolidate around.

Deliverables

  • A one-page workflow map.
  • A tools inventory spreadsheet.
  • A quick-wins list with effort and value estimates.
  • A rough target stack.

Time and budget

  • 40–80 hours of combined effort, half yours, half a partner’s.
  • €5,000–€12,000 if done with a partner, or equivalent in internal time.

Warning

The temptation at this stage is to pick tools before understanding workflows. Don’t. The wrong CRM chosen on week one is a two-year problem.

Phase 2 — Weeks 3–6: automate one workflow end-to-end

Goal

Take one full workflow — end to end, one domain — and get it running on a modern stack. This is the credibility phase. If it works, everything else gets easier politically.

Choosing the workflow

The right candidate has four properties:

  • Frequent. It happens multiple times a week, so savings compound.
  • Painful. People already hate it, so there is no change-management resistance.
  • Bounded. It touches two or three systems, not ten.
  • Visible. The owner or customer-facing staff feel the improvement directly.

Common winners: quote-to-invoice, inbound enquiry to CRM entry, time-to-invoice for services firms, compliance filing, supplier onboarding.

Work

  • Choose specific tools within the category targets from Phase 1.
  • Set up accounts, user provisioning, identity (passkeys from day one — don’t retrofit).
  • Build the workflow in whatever automation glue you settled on (n8n, Make, native integrations, custom code where needed).
  • Parallel-run for two weeks. The new way runs alongside the old way. Mistakes surface against a known-good baseline.
  • Switch over. Keep the old path cold for a further month in case of rollback.

Deliverables

  • One workflow fully migrated.
  • Documented — what it does, who owns it, what to do when it breaks.
  • Metrics baseline — how long it used to take, how long it takes now, how many errors.

Time and budget

  • Low end: €8,000–€15,000 for a simple workflow with good existing data.
  • Mid: €15,000–€35,000 for a real end-to-end migration touching multiple systems.
  • High: €35,000–€70,000 if significant data clean-up or custom integration is needed.

Warning

Scope creep kills this phase. The right answer to “can we also do X?” is almost always “yes, in phase 3.” One workflow, end to end, is a more valuable deliverable than four workflows half-done.

Phase 3 — Weeks 7–12: digitise one customer-facing channel or product

Goal

With the back office stabilised on a modern stack, put one customer-facing surface on it properly. Website, booking flow, customer portal, self-service for a repeat service — whichever matches your business.

Work

  • Design the surface from the customer’s point of view, not yours. This is the phase where “we’ve always done it this way” costs you the most.
  • Integrate with the Phase 2 back-office workflow so that customer actions flow into the real systems without re-keying.
  • Instrument it. You want to know how many people arrive, how many convert, where they drop off, what they ask for. This is cheap and transformative.
  • Soft launch. Real customers, controlled volume, fast iteration for two to three weeks.
  • Full launch with a clear rollback path and a person named as owner.

Deliverables

  • One customer-facing surface live and instrumented.
  • A dashboard the owner looks at weekly.
  • A short list of observed customer behaviours that surprised you.

Time and budget

  • Low: €10,000–€20,000 for a simple customer-facing surface on a managed platform.
  • Mid: €20,000–€50,000 for a bespoke surface integrated with your back office.
  • High: €50,000–€120,000 for a full product-shaped digital channel.

Warning

Do not skip instrumentation. A customer-facing surface without analytics is an opinion generator. With analytics, it becomes a learning system.

Honest timelines and honest failure modes

Ninety days is an aggressive but realistic pace for a business that commits to the project. Expect a couple of wobbles along the way.

What actually goes wrong

  • The data is messier than you thought. Customer lists with duplicates, product catalogues with inconsistent naming, compliance records with gaps. Budget time for clean-up; it always takes longer than people estimate.
  • Staff adoption is slower than the plan. People adopt new tools at the speed of their own confidence, not the project plan. Factor in training, office hours, and a named champion per function.
  • One vendor is the wrong choice. You will probably replace one tool chosen in Phase 1 by the end of Phase 3. That is normal. Plan for it by not over-investing in any single one until it has proved itself.
  • The quick wins get political. The hand-done compliance filing that an employee has done for 12 years is a point of pride. Be explicit that automation frees them for higher-value work, and mean it.

What “done” looks like at 90 days

  • Identity and device hygiene is reasonable.
  • One back-office workflow runs end-to-end on modern tools.
  • One customer-facing surface is live and instrumented.
  • The owner has a weekly dashboard.
  • Staff can articulate, in their own words, what changed and why.
  • There is a written plan for months 4–12.

A short working rule

Don’t try to transform everything. Pick one workflow and one surface, do them properly, stabilise, and iterate. The companies that fail at this consistently try to do eight things badly. The companies that succeed pick two and do them well.

If you are a business sitting at the start of this runway and want someone to walk the first 90 days with you, that’s the kind of engagement we take on. It is focused, honest about scope, and designed to leave you in control of what you’ve built at the end.