Back to Digital Decisions

Digital Decisions

How We Decide What Digital Work Is Actually Worth Doing

A practical prioritization system for deciding which SEO, ads, content, analytics, UX, or web development work deserves time, budget, and attention.

Gabriel Luis • May 11, 2026

Most digital strategy gets too big too fast.

The useful question is usually smaller:

What should we fix, build, test, pause, or stop doing next?

That question matters because most teams do not suffer from a lack of possible work. They have SEO recommendations, campaign ideas, tracking gaps, landing page requests, content briefs, site bugs, analytics questions, automation dreams, and half-finished experiments all competing for attention.

The work worth doing is not the work that sounds most strategic. It is the work that can change a real outcome, has enough evidence behind it, and can actually be shipped.

Start With The Constraint, Not The Channel#

Before choosing tactics, name the constraint.

Are we trying to get more qualified traffic? Reduce wasted ad spend? Improve conversion rate? Fix a technical issue before launch? Give the team better reporting? Ship a page that has been stuck in review?

If the constraint is vague, the work will be vague too.

Good constraints point to action:

  • Organic leads dropped after the migration.

  • Paid search spend is rising but qualified enquiries are flat.

  • The new service page is nearly ready, but tracking and metadata are incomplete.

  • The team has too many dashboards and still cannot tell what changed.

  • Developers need a focused fix list, not another raw audit export.

  • The contact form gets submissions, but too many are poor fit.

  • The site has useful tools, but users are missing the primary action.

Weak constraints create theater:

  • We need a better digital strategy.

  • We should do more content.

  • The website needs work.

  • We need to use AI.

  • The brand should feel more premium.

The first group points to action. The second group points to meetings.

Put Every Idea Through Four Filters#

We usually judge digital work against four filters: outcome, evidence, effort, and ownership.

FilterQuestionWhy it matters
OutcomeWhat changes if this works?Prevents activity from pretending to be value
EvidenceWhy do we think this is a problem?Keeps the team out of guesswork
EffortCan this ship in a useful slice?Avoids giant plans that never land
OwnershipWho will do the next step?Turns strategy into production

If an idea cannot pass these filters, it may still be interesting. It is not ready to be prioritized.

Decide Which Type Of Value It Creates#

Not every useful task creates revenue immediately. Some work prevents risk. Some work gives the team better information.

The point is to be honest about which kind of value you are buying.

Value typeExamplesWatch for
RevenueMore qualified traffic, better conversion, cleaner campaignsOverclaiming before measurement exists
Risk reductionLaunch QA, crawlability, broken tracking, privacy, performance, accessibilityInvisible work getting deprioritized
LearningA test that tells the team what to do nextTests with no decision attached
EfficiencyLess manual reporting, cleaner workflows, fewer support loopsSaving minutes while ignoring bigger constraints
TrustClearer pricing, proof, policies, accessibility, reliabilityTreating trust as decoration

If a task does not affect any of those, it is probably polish or preference.

Separate Evidence From Opinion#

The best tasks have evidence.

That might be Search Console data, crawl results, ad spend, conversion logs, analytics events, support tickets, page speed results, sales feedback, session recordings, or repeated user confusion.

Evidence does not need to be perfect. It just needs to be enough to avoid guessing.

SignalUseful question
Search ConsoleAre impressions, clicks, queries, or page groups changing?
Paid media dataWhere is spend leaking or conversion quality falling?
Analytics eventsCan we see the funnel step that is failing?
Crawl dataWhich technical issues affect important pages?
CRM or sales notesAre leads qualified, poor fit, delayed, or confused?
Support ticketsWhich questions should the site answer sooner?
UX reviewWhere does the task break under real constraints?

Preference is still allowed. It just should not be dressed up as evidence.

Slice Big Ideas Until Something Can Ship#

Big initiatives often hide the useful first step.

Instead of "improve SEO," ship the page title cleanup, internal link pass, schema fix, redirect check, or content refresh that can be done this week.

Instead of "improve paid media," fix the query waste, broken conversion import, landing page mismatch, or campaign structure issue.

Instead of "improve the website," ship the form fix, hero rewrite, mobile layout repair, tool empty state, or pre-launch UX check that removes friction now.

Small shipped work beats large strategic intent.

Make Ownership Boringly Clear#

If nobody owns the next step, the task is not ready.

Good work has:

  • one owner,

  • one next action,

  • a definition of done,

  • a review point,

  • a way to know whether it mattered.

The owner does not have to do every task personally. They are accountable for getting the next decision or shipment unstuck.

Separate Diagnosis From Production#

Some work is diagnostic. Some work is production.

Diagnostic work answers "what is happening?" Production work changes the thing.

Both matter, but they should not be confused.

An SEO crawl is diagnostic until someone fixes the blocked pages, missing titles, redirect chains, and canonical mismatches. A tracking audit is diagnostic until the events are corrected. A landing page review is diagnostic until copy, layout, or measurement changes ship.

The handoff between diagnosis and production is where a lot of digital work quietly dies.

Diagnostic outputProduction handoff
Crawl exportRanked fix list with page groups, owner, and impact
Analytics auditEvent changes, validation steps, and reporting owner
Landing page reviewCopy, layout, or measurement tasks ready for build
Paid search reviewNegatives, structure changes, asset tests, or bid changes
Content auditMerge, refresh, redirect, or delete decisions

Watch For Strategy Theater#

Strategy theater usually has a few tells:

  • A framework appears before the problem is clear.

  • The recommendations could apply to almost any company.

  • The work creates more categories, pillars, and diagrams than decisions.

  • Nobody can say what will ship next.

  • The team is asked to approve a roadmap before seeing the evidence.

  • Every idea survives prioritization because saying no feels uncomfortable.

Useful strategy is more direct. It reduces options. It names tradeoffs. It makes the next useful action easier to see.

Use A Plain Prioritization Table#

When a backlog gets noisy, use a table like this:

WorkEvidenceValue typeEffortOwnerNext action
Fix blocked crawl resourcesCrawler found CSS and JS blocked by robots.txtRiskSmallDeveloperUpdate robots rules
Rewrite top service page introHigh traffic, weak enquiriesRevenueMediumSEO/contentDraft new copy
Clean paid search queriesSpend wasted on poor-fit termsRevenueSmallAds leadAdd negatives
Repair checkout event trackingPurchase data incompleteLearningMediumDeveloperTest event flow
Clarify contact form success stateUsers ask whether enquiry was sentTrustSmallDesigner/devUpdate success state

The table is intentionally plain. If the decision cannot survive a plain table, it probably is not clear yet.

Decide, Defer, Delete, Or Diagnose#

The point of prioritization is not to rank everything forever. The point is to move each item into a clear bucket.

BucketUse whenNext step
DoHigh enough value, enough evidence, owner availableShip the smallest useful version
DeferUseful, but timing or capacity is wrongName the trigger for revisiting
DeleteLow value, unclear owner, or no longer relevantRemove it from the list
DiagnoseThe problem is real but not understoodRun a focused investigation

The delete bucket is important. A backlog that only grows is not a strategy. It is storage.

What Usually Goes First#

Start with work that is:

  • tied to a current business constraint,

  • backed by enough evidence,

  • small enough to ship,

  • owned by someone,

  • measurable after release,

  • hard to recover from if ignored.

That often points to practical SEO fixes, landing page changes, campaign cleanup, tracking repairs, SEO fundamentals, or small tooling improvements.

It rarely points to a giant transformation plan. Larger transformation work can be necessary, but it should still start from a real constraint and a sequence of shippable changes. That is the same principle behind the broader digital transformation guide.

The Useful Version Of Strategy#

Good digital strategy is not a document that proves everyone thought hard.

It is a way of deciding what deserves attention, what should wait, what needs more evidence, and what should be deleted from the list entirely.

When the work is chosen well, the team spends less time admiring the roadmap and more time shipping the fix.

The useful question stays wonderfully plain: what can we do next that changes a real outcome?

Further reading

Keep going from here.