How do you build a digital transformation roadmap for 2026?

What is a digital transformation roadmap?

A roadmap is the decision document that comes out of an assessment. It answers three questions for the year ahead: what will change, in what order, and how you will know it worked. It is not a list of software to buy, and it is not a vision statement. A good one fits on one or two pages, so the management team can actually use it in monthly meetings.

Most businesses already use some technology well and some badly. The roadmap is where you choose which gaps are worth closing this year, and which can wait. If you have not mapped those gaps yet, a digital maturity assessment is the usual starting point.

Why does a roadmap matter more in 2026?

Two things make 2026 different. The first is AI. McKinsey's global survey published in November 2025 found that 88% of respondents say their organisation regularly uses AI in at least one business function, but only about one-third say their companies have begun to scale their AI programmes, and 39% report any impact on enterprise-level EBIT. The same survey found that fundamentally redesigning workflows is one of the strongest factors behind meaningful business impact (McKinsey, 5 November 2025). In plain terms: using AI is common, getting value from it is not, and the difference is usually the process around it.

The second is regulation with fixed dates. On 19 November 2025 the European Commission proposed targeted amendments to the EU AI Act, known as the Digital Omnibus on AI, including a later start for some high-risk rules (European Parliament Legislative Train). Tax authorities are also moving invoicing online in stages, as the Mauritius Revenue Authority (MRA) is doing with its e-invoicing system (MRA). A roadmap that ignores known dates will be rewritten by them.

Update, 24 July 2026: the AI Act amendments were published in the Official Journal of the EU on 24 July 2026. They set fixed dates for the delayed high-risk rules: 2 December 2027 for stand-alone high-risk AI systems and 2 August 2028 for high-risk AI embedded in products (European Parliament Legislative Train).

How do you build the roadmap, step by step?

1. Start from today, not from tools. List the processes that matter most to revenue, cost and risk: quoting, ordering, invoicing, approvals, customer service, reporting. For each, note where time is lost, where errors happen and where nobody can see status.
2. Agree the business goals for the year. Two or three, in business terms: faster cash collection, fewer customer complaints, a cleaner year-end audit. Every roadmap item should serve one of them.
3. Turn gaps into candidate changes. Write each as an outcome, for example "purchase approvals take one day, not five", rather than "buy a workflow tool".
4. Score and sequence. Rate each candidate on value, effort, risk and dependency. Things that unblock others, such as clean customer and supplier data, go first. Our guide on which processes to automate first has a simple scoring grid.
5. Add the dated obligations. Put fixed external dates on the plan first, then fit the rest around them.
6. Give each item an owner, a budget and a measure. One named person, an approved amount, a first milestone within 90 days, and one number that should move.
7. Review quarterly. Drop, add or re-order items as you learn. A roadmap that never changes is usually not being used.

What does a roadmap example look like?

Here is an illustrative example for a mid-sized distribution business. It is a format, not a client case.

Quarter	Change	Owner	Measure
Q1	Clean customer, supplier and product master data; agree one source of truth	Finance manager	Duplicate records removed; invoice errors per month
Q1	Map and simplify purchase-to-pay approvals; set an approval matrix	Financial controller	Days from request to approved order
Q2	Move approvals from email into one workflow with an audit trail	Operations lead	Share of approvals inside the workflow
Q2	Check the invoicing system against the tax authority's e-invoicing requirements	Finance manager	Test invoices accepted
Q3	Pilot one AI use case on a low-risk, high-volume task, such as drafting replies to routine customer queries, with human review	Customer service lead	Handling time; error rate after review
Q4	Review access rights and key controls; refresh the roadmap for next year	General manager	Open control issues; next-year plan approved

Notice the order. Data and process come before automation, and the AI pilot comes after the workflow it depends on exists. Each line has one owner and one measure, so progress is easy to report.

Update, 1 September 2026: the MRA's e-invoicing stage for taxpayers with turnover above MUR 40 million is now listed as launched. Under section 20A(2) of the VAT Act, the Director-General can also require a business to e-invoice by written notice regardless of turnover (MRA). If your business is in scope, the Q2 e-invoicing line in this example becomes a Q1 priority. See is your business ready for e-invoicing?

What are the common roadmap mistakes?

• Too many items. Ten priorities means none. Five or six real changes a year is plenty for most mid-sized firms.
• Tools before process. Automating a confused approval chain makes the confusion faster.
• No baseline. If you do not measure today's cycle time or error rate, you cannot show improvement later. See how to measure whether transformation is paying off.
• AI as a separate project. AI works best inside a redesigned process, not bolted on beside it.
• No owner. "IT" or "the team" is not an owner.

Update, 22 July 2026: BCG's July 2026 CEO research found that nearly nine in ten CEOs see some cost or revenue benefit from AI in targeted areas, but only 14% have clearly defined the P&L impact for all AI initiatives, and high performers are roughly seven times more likely to redesign workflows end to end (BCG, 22 July 2026). That supports giving every roadmap item a measure from the start.

Who should own the roadmap?

The general manager or chief operating officer should own the roadmap as a whole, because it changes how the business runs, not only its systems. Each item then has its own owner in the business, supported by IT or an outside adviser where needed. Choices about data, such as what may be stored in which cloud service or shared with an AI tool, belong to the business and should follow its own appetite for risk.

If you want an outside view of where to start, Linkgrow's digital maturity audit ends in a prioritised roadmap in this format. You can also run the free gap finder for a first read on your gaps.

Frequently asked questions

How long should a digital transformation roadmap be?

One or two pages for the management version: five or six changes, each with an owner, a budget, a first milestone and a measure. Supporting detail can sit in separate project plans.

Should a roadmap cover one year or three?

Plan the next twelve months in detail and keep a short list of direction for the two years after. Review the detailed year every quarter.

Where should AI sit in a 2026 roadmap?

Inside the processes it supports, after the data and workflow it depends on are in order. Start with one low-risk, high-volume use case with human review and a clear measure.

Do I need an assessment before building a roadmap?

You need an honest picture of how the business runs today. A digital maturity assessment is a structured way to get one, but a careful internal review of key processes can work for a first roadmap.