Hirsh Marketing, April 2021 to present
Paid acquisition across a live agency portfolio
I own the decisions that move money on 12 to 20 Meta ad accounts at a time: what each account is forecast to deliver, how it is structured, what gets tested, what the results mean, and where budget goes next.
- My role
- Senior Media Buyer. Hands-on account responsibility.
- Platform
- Meta primarily. Google Ads through independent work.
- Working with
- Account managers, leadership, clients, and creative teams.
- Shown below
- A hypothetical budget decision, invented numbers.
01
Targeting
Who we reach, and who we deliberately do not.
02
Campaign projections
What the account should deliver for the budget, before it spends.
03
Account structure
Spend separated around the questions I will need to answer later.
04
Budget allocation
Where each dollar goes this week, and why it moved.
05
Testing decisions
One variable at a time, sized so the answer is worth acting on.
06
Pacing
Spend and volume against plan, so a miss shows up early.
07
Optimization
What to scale, change, investigate, or stop.
08
Reading the results
Platform numbers reconciled against what the business actually recorded.
Scope
- concurrent Meta ad accounts
- 12–20concurrent Meta ad accounts
- annual spend across the Hirsh portfolio
- ~$2.5Mannual spend across the Hirsh portfolio
- in the role, April 2021 to present
- 5 yearsin the role, April 2021 to present
- group coaching sessions I lead
- 2 weeklygroup coaching sessions I lead
The questions I get paid to answer.
Clients bring the offer, the copy, the creative, and the landing pages, and they know their own business better than I will. My job starts there: turning those inputs into a forecast, an account structure, a testing plan, and a recommendation with a number attached to it.
Most weeks the real work is deciding which question to ask. Is this account short of pace because delivery changed, because the offer stopped converting, or because we are counting something differently than we were last month. Those look similar in a dashboard and they get fixed in completely different places.
Then I say which one it is, and what I want to do about it, to the account manager, to leadership, and to the client. Recommending that we stop spend is part of that when the numbers call for it.
A budget decision, worked through
Hypothetical example
Seven customers is a reason to look closer, not a reason to move the budget.
Invented numbers, one simple model, every figure derived from it. The account is on target for cost and behind on volume, so the question is not whether to spend, it is whether a cheaper audience exists that the remaining budget can lean on. The test is designed to answer that one question and nothing else.
One budget decision, end to end
Hypothetical
01
The goal
120 customers at $180 or better
What the business needs this quarter, not what the account is producing. 120 times $180 sets the planned budget at $21,600 over 13 weeks.
02
Week 5 read
37 customers, $180 blended
$6,660 spent so far. Cost is on target and volume is not: 7.4 customers a week lands near 96 by week 13 if nothing changes.
03
The test
3 audiences, $1,000 each, 10 days
Creative and offer held constant, so the only thing that varies is the audience. $100 a day per cell, $3,000 in total, read on a 7-day click window.
04
After the test
53 customers, $11,940 left
67 still needed. At $180 the remaining budget buys 66. At B's $143 it buys 83. That gap is the whole decision.
05What came back, cost per customer by cell, rounded to the dollar
Audience A, Lookalike of past buyers
$2005 customers
Audience B, Broad, no interest targeting
$1437 customers
Audience C, Stacked interests
$2504 customers
06The decision
Extend A and B for 7 more days at the same budget and pause C. Move budget to B only if it holds under $160 on 15 or more customers.
Why not scale now: 7 customers is a thin read. On $1,000 of spend, 7 customers is $143 each. 8 would be $125 and 6 would be $167, so one conversion either way moves the answer, and the 7-day click window is still open on the last days of every cell. Why not stop: the prize is the difference between missing the target and clearing it, which is worth 7 days of observation at $1,400.
What a dashboard shows is a reporting artifact, not the business result.
A platform can report a cheap lead all day. Whether that lead answered the phone, showed up, or bought anything depends on what is being recorded and how. Conversion events can be defined narrowly or loosely, integrations can send downstream data or not, attribution windows differ, and data quality varies account to account.
So I reconcile platform numbers against landing page and CRM conversion data before I treat any of it as a result. When those two disagree, the disagreement is usually the most useful thing in the account.
It also changes what a good week looks like. Cost per lead falling while cost per customer rises is not a win, and if you only read the platform you will call it one.
Explaining the decision is part of the job.
I lead two weekly group advertising coaching sessions for business owners, and I have trained and onboarded other media buyers. Teaching this work to someone who does not do it daily is the fastest way to find out which parts of your own reasoning are actually solid.
That is also the part that transfers well in-house. I explain recommendations to account managers, leadership, clients, and business owners, in the terms each of them uses to make a decision.
Hiring for paid acquisition or growth?
Send me the role, the product, or the problem you are working on.