MetaMaxd
← Back to blog
Strategy

Your Ad Account's First 30 Days: What to Do When There Is No Data

MetaMaxd·November 16, 2026·6 min read

TL;DR

A new ad account has nothing: no pixel history, no audiences, no learnings. The first 30 days are not about performance. They are about giving Meta enough clean signal to learn who buys from you. Week one, get tracking right and launch one broad prospecting campaign with your best creative. Week two, fix whatever the data says is broken. Weeks three and four, add retargeting once there is actually someone to retarget, and start testing a second concept. Expect the first month to look expensive and messy. That is normal. What is not normal is launching six campaigns on day one and wondering why none of them learned anything.

Week one: tracking first, ads second

Every new advertiser makes the same mistake in the same order. They build the campaign first and check the tracking later. Do it the other way round, because Meta optimises toward the events you send it, and a new account sends it nothing until you wire it up.

Install the pixel, set up the Conversions API, and fire the full funnel: ViewContent, AddToCart, InitiateCheckout, Purchase, with value and currency on the purchase event. Then buy something from your own store, or submit your own lead form, and watch the event arrive in Events Manager. If it does not arrive, or arrives twice, fix it now, while the account is empty and nobody is watching. A duplicated purchase event in month one becomes a reporting argument in month six.

While tracking settles, pick your starting point. One campaign, one objective, sales or leads, whichever is the real business goal. One broad ad set, Advantage+ audience or broad targeting with no interests stacked on top. Three to five ads, but genuinely different concepts, not five crops of the same image. Budget enough to buy signal: if your target cost per result is 30 euros, 10 euros a day will take a week to buy a single conversion, and the system will learn nothing. Start with a budget you can afford to spend for two weeks without a return, because that is what the first two weeks are.

Week two: read the boring numbers

By week two you have some data, and most of it will look bad. Do not panic. Read it in the right order.

First, is the pixel firing cleanly? Check Event Match Quality and look for duplicates. If your reported purchases and your actual orders disagree by more than a rounding error, stop and fix the plumbing. Nothing downstream matters until this is right.

Second, are people clicking? A CTR under 0.5 percent on cold traffic usually means the creative is not stopping the scroll. That is a creative problem, and no amount of audience tinkering will fix it. A decent CTR with no conversions usually means the landing page, the offer, or the price is the problem, not the ad. The ad did its job. The page did not.

Third, resist the urge to rebuild. New accounts need stability more than they need optimisation. Every significant edit restarts learning, and an account that has never finished learning has nothing to optimise from. If an ad is truly dead, kill it. If it is merely mediocre, leave it alone until week three.

Weeks three and four: add the second layer

Now there is someone to retarget, and it is time to build the second campaign. Retargeting on day one is pointless because the pool is empty. By week three you have site visitors, video viewers, and page engagers, and a small retargeting campaign aimed at them will usually be your cheapest conversions of the month. Keep it simple: one ad set for site visitors from the last 30 days, one for video viewers, same creative with a nudge toward finishing what they started.

This is also the moment to launch your second creative concept. Not a variation, a concept. A different angle, a different format, a different opening line. The first concept taught you something about what your audience responds to. The second concept tests whether that lesson was real or a fluke. If both concepts point the same way, you have the start of a creative direction. If they disagree, you have learned that your audience is more complicated than one ad, which is also useful.

By day 30, write down what you know. Your real cost per result, not the one you hoped for. Your best creative concept and why it worked. Your landing page conversion rate. Your pixel health. That page of notes is the strategy for month two, and it is worth more than any dashboard, because dashboards show what happened and notes show what you learned.

What not to do in month one

Do not launch six campaigns. Each one needs roughly 50 optimisation events a week to exit learning, and a new account will not feed six hungry ad sets. One prospecting campaign and one small retargeting campaign is plenty.

Do not use lookalikes yet. Lookalikes need a seed audience with enough people in it, and a week of traffic is not a seed. Build the seed first, use lookalikes in month two.

Do not judge the account on ROAS in week one. There is no ROAS yet. There is only cost per learning event, and the learning is the product you are buying.

And do not copy a competitor's structure. Their account has history. Yours does not. The structure that works for them assumes data you have not collected yet.

FAQ

How much should a new Meta ad account spend in the first month?

Enough to buy signal. As a rule of thumb, budget at least 10 times your target cost per result per ad set per week, so the system gets enough conversions to learn. If you cannot afford that, narrow the scope: one campaign, one ad set, fewer ads.

How long does it take a new ad account to start performing?

Expect two to four weeks before results stabilise. The first week is tracking and learning, the second is fixing what the data exposes, and weeks three and four are when retargeting and a second creative concept start to show what the account can do.

Should a new account use Advantage+ or manual targeting?

Start broad. Advantage+ audience or broad targeting with no interest stacking gives the system the widest pool to learn from, which is what a new account needs most. Narrow targeting on day one just teaches it about a small room.

When can a new account start retargeting?

Once there is a pool worth retargeting, usually from week three. A retargeting audience built from a handful of visitors will not deliver, and it will eat budget that prospecting needs more.

What is the biggest mistake new advertisers make?

Building campaigns before the tracking works. Meta optimises toward the events it receives, so broken, duplicated, or missing events teach it the wrong lessons from day one, and everything built on top inherits the error.

Should I run lookalike audiences in the first month?

No. Lookalikes need a seed audience with real volume, and a new account has not collected one yet. Build site traffic and customer lists first, then use lookalikes in month two.

MetaMaxd is a Meta ads agency. We have launched enough new accounts to know the first month is about teaching, not earning. If yours needs a steady hand through the blind part, that is what we do.

We use cookies, including the Meta Pixel, to understand how this site is used and measure ad performance. See our Privacy Policy for details.