What to Build First: Sequencing Digital Work

Most businesses arrive with a list rather than a project. The website is dated. Organic traffic is flat. The operations team is drowning in a spreadsheet. Somebody keeps mentioning AI.

All of it is probably worth doing eventually. You can afford one thing now. The order matters more than most people expect, because the right sequence makes each project cheaper and the wrong one means doing work twice.

Sequence by constraint, not by enthusiasm

The instinct is to start with whatever is most visible or most complained about. A better question: what is currently preventing the business from growing?

Every business has one binding constraint at a time. Work that does not address it produces less than it should, however well executed.

Four common constraints, and what each points to:

Not enough people know you exist. The constraint is visibility. That points at SEO - but with a prerequisite, covered below.

People find you but do not convert. The constraint is the website. More traffic to a site that does not convert amplifies a problem rather than solving one.

You win the work but cannot deliver it efficiently. The constraint is internal. That points at automation or custom software, not at marketing.

You can deliver but cannot scale the offer. The constraint may be the product itself - which is a different conversation from any of the above.

Naming the actual constraint eliminates most of the list immediately.

The dependency that catches people out

There is one ordering rule that is close to absolute: fix the website before investing in SEO.

Not because SEO is less important, but because the two are coupled. Search work drives people to pages. If those pages are slow, unclear or unconvincing, you are paying to deliver visitors to something that will not convert them.

Worse, if the site needs rebuilding later, a rebuild that changes URLs puts the accumulated search equity at risk. You would be building authority on a structure you are about to replace.

The efficient order is: get the site right, with the search foundations built into it, then invest in ongoing search work on top of a structure that will not move.

There is a caveat. If the existing site is fundamentally sound and simply needs content and structural work, that is SEO work, not a rebuild. Not every underperforming site needs replacing - see redesign vs rebuild.

Internal work is often more urgent than it looks

There is a bias toward external-facing projects because they feel like growth. Internal tooling feels like cost.

That gets the maths wrong when delivery is the constraint. If your team spends fifteen hours a week on work that should not exist, that is capacity you are already paying for and not getting. Recovering it is usually cheaper than winning more revenue to hire against.

It is also often the fastest project on the list. A well-scoped automation removing a specific repetitive task can be built in weeks, where a site rebuild takes months. See what business processes you should automate.

The test: if you doubled your enquiries next month, could you deliver? If the honest answer is no, fixing delivery comes first.

Sequence so each project makes the next cheaper

Some orderings compound. Others create rework.

Compounds well:

• Website → SEO. The structure is right, so search work has something to build on.
• Automation → application. Automating a process teaches you how it actually runs, which is most of the specification if you later build software around it.
• Application → automation. Once a workflow lives in software, the repetitive steps inside it become visible and connectable.
• Website → application. A shared design language means the product does not look like a different company.

Creates rework:

• SEO → rebuild. Authority built on URLs you are about to change.
• Application → process design. Encoding a process nobody agreed on.
• Automation → process fix. Automating a broken workflow gives you a faster broken workflow.
• Mobile app → learning what users need. Expensive way to discover the workflow was wrong. See web app vs mobile app.

Make the first project small enough to finish

Whatever comes first, scope it so it completes.

Long projects lose momentum, absorb budget, and delay everything behind them. A six-month build that stalls at month four leaves you with nothing and a smaller budget.

Prefer a project you can finish in weeks that measurably improves something, then reassess. The constraint frequently moves once you have addressed the first one, and the second project on your original list may no longer be the right one.

This also gives you information. The first project tells you how well you work with whoever is doing it - worth knowing before committing to something larger.

A practical sequencing exercise

1. Write down the constraint. One sentence: what is currently stopping the business from growing?
2. List the candidate projects, with a rough cost and duration for each.
3. Mark which address the constraint. Set the others aside - not cancelled, just later.
4. Check dependencies. Does any of the remaining work become wasted if something else changes afterwards?
5. Pick the smallest one that moves the constraint.
6. Set a review point. After it ships, reassess. The constraint may have moved.

The discipline is step three. There is always a case for doing everything, and the businesses that get the most from their budget are usually the ones that did fewer things in a better order.

When the answer is “none of these”

Sometimes the list is all technology and the constraint is not.

If your offer is unclear, your pricing is wrong, or you have not decided who you sell to, no website will fix it. A site can only communicate a proposition; it cannot invent one. We have seen rebuilds commissioned to solve what was a positioning problem, and the new site had the same problem in a nicer typeface.

It is worth asking honestly whether the technology is the constraint before spending on it.

---

If you have a list and are not sure what comes first, describe the situation. We will tell you which project we would do, in what order, and which ones we would leave for later - including when the answer is that you do not need us yet.