Most agency relationships that go bad do not go bad in month six. They go bad in month one, quietly, and both sides spend five more months being polite about it.
I run an agency, so I have been on both ends of that. The first 30 days are the most honest signal you will ever get, because everything that will eventually go wrong is already visible if you know what to look at. This is what a real first month looks like, week by week, and what it looks like when it is going sideways.
What should happen in the first 30 days with a social media agency? Week one is an audit of your existing channels and a strategy that visibly came out of it. Week two is production, usually a single shoot day that banks a month of content at once. Week three is first publish, plus an early read on which formats are landing. Week four is the first real report, tied to business outcomes rather than likes. By day 30 you should have content running, a named person accountable for it, and numbers you understand. The two most common failure modes are an agency that never produces anything original, and a client who cannot get access or approvals turned around fast enough to let production start. Both are visible inside two weeks, and both are fixable if somebody says so early.
Week one: the audit nobody sends you
The first week is not content. It is diagnosis.
A real audit looks at what you have already published and says specific, slightly uncomfortable things about it. Which posts actually performed and why. Which format you keep making that nobody watches. Where your competitors are showing up that you are not. Whether your profile converts a visitor into anything at all.
Then the strategy comes out of that audit. This is the part to watch. If the strategy could have been written before anyone looked at your accounts, nobody looked at your accounts. A generic pillar framework with your logo on it is a template, not a plan.
What you should be able to ask at the end of week one:
- Which of my existing content are you keeping, and which are you killing?
- What are we posting, on which platforms, how often?
- What are we trying to make happen, in business terms?
If those get answered with specifics, week one worked. If they get answered with adjectives, push back now rather than in month three.
What you owe in week one: admin access on every account, not a shared login, and one person who can approve content without convening a committee. Slow access is the single most common reason a first month drags, and it is almost always on the client side. We have had onboarding stall for eleven days waiting on an Instagram permission.
Week two: the shoot
This is where most of the difference between agencies shows up.
An agency that produces original content will book a shoot day in week two. One day, one setup, and a month or more of assets out of it. That is content batching, and it is the only way the economics of consistent video work at a retainer price. Shooting weekly costs a fortune and burns out everyone involved.
An agency that does not produce will instead ask you to send them content. That is a real business model and it is cheaper, but you need to know that is what you bought before month two, not after.
A batched shoot day usually runs four to six hours and covers:
- Talking-head pieces answering the questions your buyers actually ask
- Product or process footage, whatever your version of showing the work is
- Behind the scenes, which reliably outperforms the polished cut
- Enough b-roll to cut everything else together for a month
What you owe in week two: yourself, or whoever on your team is going on camera. This is the one thing that cannot be handed back to the agency, and it is the thing clients most often try to reschedule. Every week that shoot slips, the whole calendar slips behind it.
If your team hates being on camera, say so in week one. There are ways to work around it, and finding out on the shoot morning is not one of them.
Week three: first publish, and what the early numbers actually mean
Content starts going out. Now the temptation is to read the numbers, and this is where a lot of good engagements get judged badly.
Three weeks in, the honest read is:
What is worth looking at. Are people watching past the first three seconds? Is anything getting saved or shared, which matters more than likes because it signals the content was useful. Are profile visits climbing.
What is not worth looking at yet. Follower count, which lags everything. Leads, because the sample is too small. Revenue, obviously.
The useful question in week three is not "is this working." It is "which of these is working," so the next batch can be weighted toward it. A good agency is already narrowing by now, and will tell you which format they are doubling down on and which one they are dropping.
You should also be able to see what is publishing next week without asking twice. If your content calendar is a mystery, that is a process problem that gets worse, not better.
Week four: the first report
The first report is the clearest tell in the entire month, because it shows you what the agency thinks its job is.
A weak report is a screenshot of platform analytics. Reach, impressions, follower growth, an up arrow, no interpretation. It is technically true and tells you nothing you could act on.
A real report connects the work to the business:
- What we published and why those choices
- What performed, what did not, and the read on why
- What that changes about next month
- Where the leads or inquiries came from, and honestly which ones we cannot attribute
That last part matters more than people expect. Social attribution is genuinely messy, and an agency that pretends otherwise is either naive or managing you. Being straight about what cannot be tracked is a trust signal, not a weakness.
For context on what the far side of this looks like when it works: across our social clients, average reach increase runs about 600 percent, and consistent managed social generated $120K+ in new business for Dwight DeLoach. Neither of those happened in month one. Month one is where the machine gets built.
When month one is going wrong
Not every rough start is a bad engagement. Some of these are fixable if you name them early. But two or more together, inside 30 days, is a real signal.
Nothing original has been produced by the end of week two. The most reliable tell there is. If the calendar is filling with stock imagery and reposts, you bought scheduling, not production.
You are still the one supplying content. Fine if that is the deal you signed. A problem if it is not.
Your point of contact keeps changing. Usually means your account is being passed around, and every handoff resets the context you paid to build.
Nothing has been revised after feedback. You gave notes in week two and week four looks identical. Either the notes did not reach the person making the work, or nobody is accountable for incorporating them.
Reporting stops at engagement. Covered above, and it is the tell that predicts the most frustration at month six.
You cannot get a straight answer about next week. A calendar that only exists in someone's head is a calendar that will slip.
If you see these, say it in month one, plainly. Good agencies fix it. The ones that get defensive about a direct question in week four are showing you month six early, which is useful information.
If you have not signed yet
Three questions worth asking before money changes hands. They cut through more than a portfolio review does.
Who produces the content, and did you make this? Point at a specific piece of their work and ask who shot it, who wrote it, who edited it. A production-capable team answers in detail immediately, because they were there. A reseller gets vague or credits a partner. Neither is disqualifying. Not knowing which one you hired is.
Who is on my account, and how many others do they carry? You want a name, a role, and a number. The number is the part nobody volunteers. Someone carrying eight accounts is stretched. Someone carrying three is present.
What will you report, and how does it tie to money? Ask what a monthly report contains, then ask which of those numbers connects to revenue. The answer tells you what they think they are being paid for.
One more thing worth reading before the deliverables list: the exit terms. Notice period, who owns the content and raw footage when it ends, and whether you keep full admin on your accounts. Get all three in writing. If leaving is clean, staying is a choice, and that changes the whole relationship in your favour.
The short version
Week one is an audit and a strategy that came from it. Week two is a shoot. Week three is first publish and an early read on format. Week four is a report that talks about your business, not your impressions.
By day 30 you should have original content running, one accountable person, and numbers you actually understand. If two of those three are missing, say so that week.
For what it is worth on our side: our social media marketing retainers run $5,000, $8,500, and $12,000 a month depending on content volume, how much video is included, and how many channels are live. The first 30 days look the same at every tier. Only the volume changes.
If you want to see what month one would look like for your business specifically, tell us what you are working with and we will map it out.