Industry
Part of Campaign management tools: what they take over, step by step
Campaign management mistakes at influencer agencies, and who makes them
Campaign management mistakes at influencer agencies have owners: what buyers, teams, clients and vendors each get wrong, and how each one shows up.
Software rollouts in influencer agencies rarely fail in one dramatic way. They fail in four smaller ways at once, each with a different author: the buyer, the team, the client and the vendor.
Sorting campaign management mistakes by who makes them is more useful than sorting by stage. It shows who must change something to stop the problem. Nobody on this list behaves unreasonably. Every mistake here is a locally sensible decision with a bad total, and in an influencer agency the total lands on creator records, briefs, approvals and disclosure checks.
What to take away
- Most rollout failures are process failures that a purchase made visible, not caused.
- The team's version of the mistake is always a speed problem. Fix the friction, not the compliance.
- A client's data expectations belong in the tool decision, because you cannot renegotiate them later.
- Disclosure monitoring is part of the record, not an extra. If the system cannot show it, the agency cannot either.
The buyer's mistakes
- Buying the demo's problem instead of yours. The demo was built around a bottleneck the vendor is good at, such as creator discovery. Yours may be briefing, approvals or creator payments, and after an hour of watching, the demo's problem feels like it was there all along.
- Choosing on breadth because breadth is easy to compare. Feature counts are commensurable and outcomes are not, so the comparison drifts to whatever can be put in a column.
- Treating price as the cost. Migration, double running, training and the month where nothing is trusted are all real and none of them is on the quote.
- Skipping the export test. Pull out a creator list, a campaign's approvals and its posting history before you sign. It is the cheapest thing in the evaluation and the only one that limits the damage of being wrong.
- Refusing to stop. Once a few weeks have gone into a comparison, walking away starts to feel like waste rather than a result, which is sunk cost doing the deciding.
The last one is worth naming out loud, because it turns a bad shortlist into a bad purchase. The wider method for keeping a comparison honest is set out in the guide to comparing suppliers.
The team's mistakes
The team is not resisting the tool. Account managers and creator leads are under time pressure, and every one of these is a shortcut that pays off today.
Team shortcuts and their costs
Shortcut
- Personal spreadsheet
- Faster, no required fields
- Batch logging
- Client call comes first
- Notes field for all
- Structured field does not fit
- Shared login
- Access takes two days
- Approve in chat
- Client is in chat
Why it is rational
- Personal spreadsheet
- Two records, no truth
- Batch logging
- Accurate only on Fridays
- Notes field for all
- Facts nobody can search
- Shared login
- Audit trail says nothing
- Approve in chat
- Approvals in one history
Cost in three months
- Personal spreadsheet
- Batch logging
- Notes field for all
- Shared login
- Approve in chat
| What people do | Why it is rational for them | What it costs three months later |
|---|---|---|
| Keep the personal spreadsheet running | It is faster and it never asks for a required field | Two records, and no way to tell which one is right |
| Log work in batches at the end of the week | The client call is now and the logging is not | A system that is accurate on Fridays and wrong the rest of the time |
| Use the notes field for everything | The structured field does not fit the situation | Facts nobody can search, report on or hand over |
| Share one login | Getting someone access takes two days | An audit trail that says nothing about who did what |
| Approve in chat and record it later | The client is in chat | Approvals that exist in one person's history |
The pattern holds in all five rows: the shortcut is faster now, and the cost lands on someone else later. Asking people to comply does not fix it, but fixing the two seconds of friction usually does.
The shortcut that costs most in influencer work is the last row. A creator's post approved in chat, with the disclosure check done by eye and never logged, leaves no record that anyone checked it.
Where the tool genuinely cannot fit the work, the honest answer is to keep that part outside it and say so. Do not record a fiction.
The client's mistakes, which become yours
Clients do not make tooling mistakes on purpose. They make commitments that only bind you.
A client asks for reporting in their own system's format, so every campaign carries a manual export step forever.
A client requires their data not sit with a third party, and that requirement arrives after you have signed.
A client asks for their workspace back at the end of a retainer, having never been told what a workspace contains: creator contacts, briefs, contracts, content files and approvals.
A client expects proof that the creators were told how to disclose and that someone checked. The Federal Trade Commission says that advertisers need reasonable programs to train and monitor members of their network, and that a brand should ask for regular reports confirming the program is operating properly. When an agency runs the creators, those reports come out of its campaign records or not at all.
None of these are unreasonable requests. All of them are much cheaper to answer before a purchase than after one.
The habit that prevents the whole class is cheap. Work out which parts of the workflow in the guide to what these tools actually do touch client records at all, then ask your two largest clients what their data, disclosure and reporting requirements are, in writing, before you shortlist. It takes a week and it removes candidates.
Where client records live and what happens when the relationship ends is not only a commercial question. For agencies with British clients, the UK regulator's guidance on data protection responsibilities is a reminder that duties attached to processing someone else's records do not transfer with the subscription.
The mistakes a vendor is happy to let you make
Nothing dishonest is required for these. A vendor simply has no reason to interrupt you.
Vendor-friendly mistakes to avoid
- Configuring around a process about to change
- Buying seats for twice-a-year users
- Accepting a pilot on sample data
- Letting whoever is free implement
- Turning on every module at once
- Buying seats for everyone on day one. Seats nobody uses still renew, and they make adoption look worse than it is.
- Accepting the default roles. Out of the box, permissions rarely match who may see a creator's rates or a client's contract.
- Leaving the renewal terms unread. Notice periods and renewal dates are in the order form, not the demo.
- Never asking how data leaves. The export test belongs before signature; after it, the answer is the vendor's to give.
- Turning everything on at once.
The order an agency should buy and switch things on is worked through in the guide to what to buy first.
Turning everything on at once has the cheapest fix of the five: sequence it. One workflow, one team, one month, then the next. Start with the workflow that carries the most risk if it goes wrong, which for most influencer agencies is content approval.
Where the mistake is not a mistake
Some failures are correct. Contracts stay outside the campaign tool because the legal team's system is the one that counts, and a client stays on a spreadsheet because four campaigns a year would never repay the setup cost.
Refusing to move a workflow in is correct, too, because the tool would record it worse than the current mess does.
The test is whether the exception is written down and owned. An exception with a name against it is a decision. An exception nobody has recorded is shadow work that will be discovered by the person who inherits the account.
Writing those exceptions down at cutover, rather than meeting them a year later, is part of the sequence described in the guide to putting a tool into service.
Common questions
Adoption stalled after a month. Is the product wrong?
Possibly, but check the friction first. Most stalls trace to a step that takes too long at the wrong moment. That is fixable without a migration, and a migration will not fix it if you carry the step across.
Should we force everyone off spreadsheets?
Not by instruction. Find out what the spreadsheet does that the tool does not, because the answer is usually something specific and small. Fix that, and the spreadsheet dies on its own.
Who should own the system after launch?
Someone who does the work it records, with time formally allocated. Ownership by a manager who never opens it produces clean permissions and rotten data.
We picked wrong. Do we switch or stay?
Neither, until you can say what specifically is wrong and whether the next product fixes that thing. Switching on general dissatisfaction usually reproduces the problem with a new interface and a new migration bill.
