I Worked Inside an Influencer Marketing Agency for 8 Months — Here’s What Clients Never See

Personal experience · Influencer marketing · Brand strategy

• My first week at the agency, I was handed a spreadsheet of 47 “micro-influencers” and told to reach out to all of them by Thursday.

• No brief. No context about the campaign. Just names, Instagram handles, and a rough budget range. My manager dropped by my desk for about 90 seconds to explain what we were selling — then disappeared into back-to-back meetings for the rest of the day.

I sat there thinking: this is how millions of dollars get spent?

The answer, over the following 8 months, turned out to be: yes. Mostly. With a few eye-opening exceptions.

• I’m not writing this to trash the industry. Influencer marketing genuinely works when it’s done right — I watched it happen. But there’s a pretty enormous gap between what clients think they’re buying and what’s actually happening on our side of the table. Brands spending real money deserve to understand that gap before signing anything.

The “strategy deck” is often built backward

When a new client signed on, one of the first deliverables was a big strategy deck. 30–40 slides, professionally designed, full of audience insights, campaign objectives, and influencer “profiles.” It looked incredibly thorough.

What I didn’t realize at first was that the deck was frequently built after informal decisions had already been made — based on which influencers the agency already had relationships with.

The process usually went like this:

1.Account manager mentally shortlists influencers they’ve already worked with (or ones their contacts represent)

2.Strategy deck gets built to support those choices, not discover new ones

3.Client receives deck and believes the selection came from rigorous, data-driven analysis

4.Data from tools like Modash or Heepsy gets pulled and dropped into slides — often that same morning

• This isn’t always cynical. Existing relationships genuinely speed up execution, and speed matters in campaign work. But clients should know that “strategic fit” sometimes translates to “we’ve worked with this person before and she replies to our emails.”

“We’re not picking these creators based on your audience — we’re picking them based on ours.”
Something no agency will ever say out loud, but is often quietly true.

Follower count is the last thing that should matter — but it’s the first thing clients ask about

• Every single client kickoff call had the same moment. Someone on the brand side — usually not the marketing lead, but a founder or a senior exec dialing in — would ask: “So how big are these influencers we’re working with?”

And you could visibly see the account managers recalibrate their answer depending on who was asking.

• Because the honest answer is complicated. Follower count, on its own, means almost nothing. An account with 800,000 followers and a 0.4% engagement rate is worth far less than an account with 40,000 followers and a 6% engagement rate in the right niche.

Large account, low engagement

800K followers · 0.4% ER

Smaller account, high engagement

40K followers · 6.1% ER

The metrics that actually drive results — saves, shares, story replies, link-in-bio clicks, conversion data — are harder to screenshot for a client report. So what got reported instead? Reach and impressions. Which are, essentially, vanity metrics wearing a business suit.

What you should actually ask before signing a campaign

1.What’s the average engagement rate on the influencer’s last 12 posts — not their profile average, which can be inflated by viral outliers?

2.What does the audience breakdown look like? Age, location, gender — pulled from a platform like HypeAuditor, not estimated by the creator.

3.What’s the audience authenticity score? This flags fake followers and bot engagement before you’ve spent a cent.

4.Can we see past campaign results? Not impressions — actual click-throughs or conversions from a similar brand.

Influencer rates are almost entirely made up

I used to think there was some kind of standard pricing formula in this industry. There isn’t.

• Rates vary wildly based on the agency’s relationship with the influencer, whether that influencer is repped by a talent manager (who adds their own margin), what the brand’s perceived budget is, and honestly — sometimes just what the influencer felt like charging that week.

• I watched an agency negotiate a creator’s fee from $4,200 down to $1,800 in a single email chain. I also watched a different team pay $6,000 for a story series that a nearly identical account did for $900 two months earlier — because nobody bothered to benchmark.

Common mistake

• Clients assume the agency negotiates hard on their behalf. Often, maintaining the relationship with the creator matters more to the agency’s long-term business than saving the client a few hundred dollars per post.

• The tools we used to sanity-check rates were pretty simple: a pricing benchmark spreadsheet we kept internally, plus occasional checks against the Creator.co marketplace and the Influencer Marketing Hub’s rate calculator. Neither is perfect, but having any reference point is better than just accepting the first number you hear.

The content approval process is where campaigns actually die

This one caught me completely off guard.

• Most campaign delays didn’t come from slow influencers or late briefs. They came from the client-side approval process. A creator would submit their draft content on time. Then it would sit in a shared Google Drive folder for 11 days while someone waited for sign-off from legal.

• By the time it got approved — sometimes with a list of 15 revision requests — the creator had moved on mentally, the timing window for the campaign had shifted, and the whole thing felt rushed and off.

• The best campaigns I worked on had one thing in common: a single client-side decision maker with actual authority to approve content. Not a committee. Not a chain of VPs. One person who could say yes or no within 48 hours.

What works

• Before a campaign launches, agree in writing on a maximum revision count (2 rounds is standard), a response window (48 hours is reasonable), and clear guidelines on what “brand safe” actually means for your company. Vague briefs create endless revision cycles.

The real tools running most campaigns (that clients rarely hear about)

Here’s a quick look at what was actually open on our screens every day — not what was listed in the pitch deck.

Modash: Influencer discovery and audience analytics

HypeAuditor: Fake follower detection, authenticity scores

Notion: Campaign briefs, internal tracking, timelines

Later: Scheduling and post performance tracking

Sprout Social: Post-campaign reporting and benchmarks

Slack + Google Sheets: Where all the real decisions actually happened

The point isn’t that these are secret or bad. It’s that clients paying premium retainers sometimes imagine a more sophisticated backend than a shared spreadsheet and a Slack channel. Knowing what’s actually being used helps you ask smarter questions about how their process actually works.

Reporting numbers are optimized to look good, not to tell the truth

This is the one I feel most uncomfortable writing, but it’s also the most important.

• End-of-campaign reports are almost always structured to highlight the wins and bury the underperformers. If a campaign had 10 influencers and 3 of them knocked it out of the park, those 3 would lead every slide. The other 7? Mentioned briefly, averaged in, or simply excluded from the “highlight” section.

• I once sat in a reporting call where we presented a campaign as having “exceeded reach targets by 34%.” Technically true — total impressions were up. But two of the influencers in the mix had clearly purchased fake engagement, something we knew from our own HypeAuditor pull but never flagged to the client. Those inflated numbers were baked into the success story.

What I’d suggest doing instead:

1 .Request the raw data export, not just the designed report. Most platforms let agencies export this — ask for it.

2. Ask to see per-creator performance broken out individually, not blended averages.

3. Compare the engagement rate promised in the strategy deck versus what was actually delivered per post.

4. For e-commerce brands, unique discount codes or UTM links per creator are the only reliable way to tie revenue back to a specific influencer. If this wasn’t set up before launch, push for it next time.

The influencer relationship stuff nobody talks about

Here’s something softer but genuinely important.

• The best creators on our roster — the ones who consistently produced content that actually moved the needle — were the ones who felt like collaborators, not vendors. When an account manager took 10 minutes to actually talk to a creator about their audience before sending a brief, the resulting content was almost always better.

• The worst results came from campaigns where we sent a 12-page PDF brief, told the creator exactly what to say, and then complained when the content felt stiff and unnatural. Which, obviously, it was.

• Audiences can tell when an influencer is reading from a script. It’s not subtle. The creator’s genuine voice is the entire point of the format — if you’re suppressing it with an overly rigid brief, you’ve basically just bought awkward advertising at influencer prices.

• The brief should answer: who is this for, what problem does it solve, and what feeling do we want viewers to have? It should not specify exact wording, required camera angles, or the number of times the product must appear on screen.

What I’d tell a brand spending money on this for the first time

• If you’re a founder or marketing manager about to sign your first influencer campaign contract, here’s the actual checklist I wish existed when I was sitting on the agency side watching money get spent:

Ask for a discovery breakdown — how were these specific creators selected? What criteria, and what data was used?

2. Set up tracking before the campaign starts — UTM links, discount codes, affiliate URLs. If it’s not trackable, you won’t know what worked.

3. Define success metrics upfront, in writing, before any content goes live. “Good reach” isn’t a metric.

4. Request an HypeAuditor or Modash pull on every creator before confirming the roster. Any agency worth working with should already have this — ask to see it.

5. Start small — one or two creators with clear tracking, then scale what works. Don’t dump your entire quarter’s budget into an untested campaign.

6. Read the raw report data yourself. Don’t just look at the designed slide deck. Ask for the spreadsheet.

Start small — one or two creators with clear tracking, then scale what works. Don’t dump your entire quarter’s budget into an untested campaign.

Read the raw report data yourself. Don’t just look at the designed slide deck. Ask for the spreadsheet.

---

One thing that genuinely surprised me

For all my complaints, I watched influencer marketing work in ways that felt almost magical when it came together right.

• There was a small skincare brand — maybe $2M in annual revenue at the time — that ran a campaign with four mid-tier creators (none over 80K followers) in a really specific niche. Total spend was around $12,000 including agency fees. Within three weeks they’d sold out two SKUs and added about 4,000 new email subscribers.

• The difference wasn’t budget. It wasn’t follower count. It was that the creators actually used the product beforehand, actually liked it, and were given genuine freedom to talk about it in their own way. The brief was two paragraphs. The content felt real.

That’s still the formula. It’s not complicated. It’s just not always what agencies are incentivized to deliver.

• So if you’re going into this space — as a client, a creator, or someone just starting in the industry — go in knowing that the gap between the pitch and the reality is wide. Not always maliciously wide. Just… wide. And knowing that going in means you can ask better questions, set better expectations, and hopefully end up on the side of the story where it actually worked.

I started my own influencer marketing agency and it’s doing …

I Tested 9 Freelancing Platforms for 6 Months—Here’s Which One Actually Paid