Scope · Jul 31, 2026 · 10 min read

Scope creep examples, and what each one costs.

Eight requests that arrive on ordinary client projects, written the way they actually reach you, with the hours and the dollars attached to each one.

Scope creep almost never arrives as one big demand you could reasonably push back on. It arrives as an extra round of revisions, a new stakeholder with opinions about work you already signed off, a mobile version nobody wrote down, and a dozen small asks that each felt too minor to bill for. The eight examples below are the ones we see most often in agency work, priced at a $125 hourly rate so the arithmetic is visible. On a single $45,000 site build they come to 93 unbilled hours and $11,625.

Most articles on this topic are written for project managers inside large organizations, where the person adding work is a colleague rather than the client paying the invoice, which changes the problem completely. An internal stakeholder can be told the change board meets on Thursday, while a client who is mid-project and mid-relationship gets a much softer answer, and the softer answer is where the money goes. Everything here assumes the person asking is the one signing the checks.

What counts as scope creep and what is just a normal change?

Scope creep is extra work that enters a project without anyone changing the price, the deadline, or the written agreement. The same request, priced and approved in writing before the work starts, is a change order, and a change order is not a failure of any kind. It is the mechanism doing exactly what it exists to do. The work is identical in both cases, which is worth sitting with for a second, because it means the only thing separating a paid change from an unpaid one is whether the request stopped somewhere before it reached a keyboard.

That distinction also explains why the examples below are so hard to catch. None of them look like a dispute at the moment they arrive. They look like a friendly message from someone you like working with, and the cost only becomes visible weeks later when the project is running two rounds long and the margin has quietly gone. If your paperwork is vague about the boundary, and the difference between a scope of work and a statement of work is a good place to start tightening it, every one of these lands as a favor rather than a purchase.

What are the most common scope creep examples on a client project?

The most common ones show up in the first half of the project, sound entirely reasonable, and cost between a thousand and two thousand dollars each. They tend to be requests that a client genuinely believes are already included, which is why arguing about them after the fact rarely goes well.

"Can we see one more version of the homepage?" The agreement said three rounds of revisions and you are on round four, but the client is being nice about it and the work is nearly done. Six hours of design plus two hours of handling the feedback comes to eight hours, or $1,000. This one is dangerous because it sets the precedent for every remaining page, and the fifth round is much easier to ask for than the fourth was.

"Can you just make sure it works on mobile too?" The scope described desktop templates and everyone assumed the rest would follow, because to a client a website is a website. Fourteen hours of layout work, testing, and fixing the two components that break below 480 pixels comes to $1,750. The word "just" is doing an enormous amount of work in that sentence, and it appears in almost every request of this shape.

"Our new head of marketing has some thoughts." Someone arrives at week six who was not in the kickoff, has no attachment to the decisions already approved, and reasonably wants to make their mark. Eleven hours of rework on approved screens plus three hours of catch-up calls comes to fourteen hours, or $1,750. The real damage is that the approvals you collected are now being treated as opinions rather than decisions.

"While you are in there, could you also…" This is the drive-by, attached to the end of a message about something else entirely. Any single one costs about forty minutes and nobody would dream of invoicing for it. Across a three-month project we routinely count nineteen of them, which is roughly thirteen hours, or $1,625, spread so thinly that it never once looked like a decision anyone made.

Which scope creep examples do the most damage late in the project?

The late ones cost more because they land on a schedule that is already committed, and because by that point nobody wants to be the person who slows down a launch. They are also the requests where the gap between what the client imagines and what the work actually involves is widest.

"We would like to launch with real content instead of the placeholder text." Content entry sits in a strange place where clients assume the agency does it and agencies assume the client will. Twenty-two pages at roughly twenty-five minutes each, including formatting, image sizing, and the inevitable second pass once the client sees it live, comes to nine hours, or $1,125. Nobody scoped it because nobody thought of it as work.

"Can you add the booking integration? It is just an embed." The embed takes four minutes and everything around it takes three days. Styling it so it does not look like a different website, handling the states where the third-party service is slow or down, testing the confirmation emails, and mapping the fields to what the client actually collects comes to twenty-four hours, or $3,000. This is the single most expensive example on the list, and it is the one most often described as small.

"Legal needs to review this before we go live." A two-week hold costs you differently from the others, because the hours are not the main loss. Your team has a slot booked for this project and now has to fill it, then unfill it, and the final invoice moves a month to the right. Call it five hours of replanning at $625, plus a cash flow gap that a small agency feels more sharply than the fee itself.

"Could you send over the source files and walk our team through it?" Handover is the part of the project that exists in everyone's head and in nobody's agreement. Organizing files, writing a short reference document, and running a training call with three people who each have different questions comes to six hours, or $750. It arrives after the last invoice has usually been sent, which is what makes it awkward to price.

The eight, totaled Extra revision round $1,000, mobile version $1,750, new stakeholder $1,750, nineteen drive-by asks $1,625, content entry $1,125, booking integration $3,000, legal hold $625, and handover plus training $750. That comes to ninety-three hours and $11,625 on a single project, none of which anyone quoted.

What does all of that add up to on one project?

On a $45,000 build, $11,625 of unbilled work is just under twenty-six percent of the fee, and it lands entirely on the margin rather than being spread across the whole number. An agency working at a thirty percent margin has about $13,500 of profit in that project before any of this happens, so absorbing all eight examples leaves roughly $1,875 for three months of work by several people. That is the arithmetic that makes an otherwise healthy-looking year feel strangely tight, and we have written separately about what scope creep costs an agency across a full year of projects.

The number is deliberately conservative in two places. It assumes a $125 rate, which is below what many agencies charge, and it counts only the hours that produced work. It excludes the meetings about whether the work was included, the internal conversations about whether to raise it, and the mental cost of a team watching the same pattern repeat on the next project.

Ninety-three hours nobody quoted, on a project that was priced for none of them.

Which of these should you absorb and which should you price?

Absorb anything under an hour that does not repeat and does not change a deliverable, and price everything else. We would put the line at roughly two hours, or $250 at a $125 rate, which is low enough that the genuinely trivial favors still happen and high enough that the mobile version and the booking integration never quietly become free. Below that line the friction of raising it costs more than the work does, and clients notice generosity at that scale in a way that is good for the relationship.

Setting the threshold matters much less than catching the requests that slip under it. The drive-by asks are the ones that break the rule, because each is comfortably below the line and nineteen of them are $1,625. The fix is to price them as a group rather than individually, which means keeping a running note and raising it once the total passes your threshold. "We have picked up about nine hours of small additions since we started, so I want to flag it before it grows" is a sentence a reasonable client accepts. The same conversation about a single forty-minute favor makes you look difficult.

There is one case where pricing is the wrong call even above the threshold. If the ambiguity is genuinely yours, meaning your scope could honestly be read the way the client read it, absorb the first instance and fix the wording for the next project. Charging a client for your own vague sentence is a fight you might win once and will pay for later, and handling scope creep without losing the client depends on being able to tell those two situations apart.

How do you stop the same examples from repeating next time?

You stop them by naming them in the agreement before they happen, using the exact language of the examples above rather than a general clause about additional work. A priced exclusion that says "mobile and tablet layouts are not included and can be added for $1,750" does more for you than three paragraphs of change-control policy, because it turns an awkward refusal in month two into an ordinary purchasing decision in month zero. Write one for revisions beyond the agreed rounds, one for content entry, one for third-party integrations, and one for handover and training, and you have covered five of the eight examples on this page.

The second piece is deciding, out loud and early, who is allowed to approve a change and what happens when someone new joins the project. The kickoff conversation that names the approver is where that gets settled, and it takes about four minutes. The third piece is making the approval itself easy enough that using it is less effort than skipping it, which usually means a short change request form that states the work, the price, and the effect on the delivery date, and a written yes from the named approver before anyone opens a file.

None of this requires becoming the agency that bills for every email. It requires the boundary being visible enough that both sides can see it, so that the moment a request crosses it, saying so feels procedural rather than personal. That is the whole difference between an agency that ends the year at the margin it planned and one that cannot quite explain where the money went.

Frequently asked questions

What are the most common scope creep examples in agency work?

The extra revision round, the mobile version that was never scoped, a new stakeholder arriving mid-project and reopening approved work, and the steady drip of small asks attached to unrelated messages. Content entry, an integration described as a simple embed, a legal or compliance hold, and an unscoped handover session round out the list. Every one of them sounds reasonable in isolation, which is exactly why they get absorbed instead of priced.

Are all scope creep examples the client's fault?

No, and assuming so will cost you money. A large share of it comes from the agency side, either because the scope was written vaguely enough to be read two ways, or because nobody wanted to interrupt a friendly relationship with a price. Clients rarely know what falls outside the agreement unless the agreement says so in plain language with numbers next to it.

What's the difference between scope creep and a change order?

A change order is the same extra work with a price, a delivery impact, and a written yes from the person allowed to approve it. Scope creep is that work without any of those three things. The work itself is identical, so the only variable that decides whether you get paid is whether the request stopped for approval before anyone started building.

How much scope creep is normal on a fixed-fee project?

Most agencies build a small allowance into their pricing, often somewhere between three and five percent of the fee, to cover the genuinely minor asks that aren't worth the friction of a change order. The trouble starts when unpriced additions run well past that allowance, because a fixed fee has no mechanism for absorbing twenty percent of extra work without eating the entire margin.

Can you charge for extra work after you've already done it?

You can raise it, but you'll usually lose the argument, because a client who has already received the work has no reason left to agree to a price for it. The realistic recovery is to stop absorbing anything further, put the remaining requests through a written approval, and treat the delivered work as the cost of learning where the boundary was. Retroactive invoices damage relationships far more often than they collect.

How do you price a scope change once the project is already running?

Estimate the hours honestly, multiply by your standard rate rather than a discounted project rate, and add the effect on the delivery date as a separate line. Mid-project work costs more than the same work quoted up front because it interrupts a schedule that was already planned, and a change that adds three business days should say so in writing rather than quietly absorbing the delay.