Optimising Meta Ads: Budgets, Bidding, A/B Tests, Ad Fatigue and the Monthly Report

Edukatz Meta TeamUpdated · 4 min read · 778 words
BUDGET & OPTIMISATIONOptimising Meta Ads:Budgets, Bidding, A/BTests, Ad Fatigue and…Edukatz Meta · Meta Ads guide
Quick answer

Plan the budget backwards: leads needed × cost per lead = monthly budget. Start with the Highest volume bid strategy, test one variable at a time, and check results weekly. Frequency above about 3 with falling CTR means ad fatigue — add new creatives. Scale winners by 20–30% every few days. Report spend, reach, results, cost per result, best creatives and next steps every month.

Launching a campaign is the easy part. The real work — and what clients pay for — is the weekly routine of reading numbers, making small changes and explaining results. Here is that routine, from budget to monthly report.

1. Plan the budget backwards

Start with the business goal, not a random amount:

  • Leads needed per month: 120
  • Expected cost per lead: ₹250
  • Monthly budget = 120 × ₹250 = ₹30,000 → about ₹1,000 a day
  • If 10% of leads buy a ₹1,999 course: 12 sales = ₹23,988 revenue → ROAS 0.8×. Not profitable — so either the close rate, the price, or the cost per lead must improve before scaling.

Also check structure: 120 leads a month is about 28 a week. That's below the ~50 results an ad set needs to leave learning — so use one ad set with a broad audience rather than three small ones.

2. Choose a bid strategy

  • Highest volume — the default. Gets the most results for the budget. Start here.
  • Cost per result goal — aims to keep the average cost near your target. Use after you have steady results and know a realistic target.
  • Bid cap — limits the bid in each auction. For advanced users; set it too low and nothing delivers.
  • ROAS goal — for sales campaigns with purchase values tracked.

3. Test one thing at a time

An A/B test compares two versions with one difference: a Reel vs an image, Advantage+ audience vs interests, Higher-intent vs More-volume form. Give both the same budget, run at least 7 days and aim for 20+ results in total before you call a winner. After the test, turn off the loser and move budget to the winner — then test the next idea.

4. Read the numbers in order

MetricIf it's bad, look at…
Cost per resultThe final number. Then find which of the metrics below caused it.
CPMAudience too narrow, high competition season, low ad quality.
CTR (link)Creative and hook; offer not interesting.
Result rate (results ÷ clicks)Landing page, form length, WhatsApp welcome message, slow replies.
FrequencyAbove ~3 a week with falling CTR = ad fatigue.

Use breakdowns (age, gender, placement) to find where money is wasted — for example the Audience Network bringing cheap clicks but no leads, or 55+ ages with a cost per lead three times higher.

5. Fix ad fatigue

When frequency climbs and CTR drops week after week, add new creatives (a new hook or format), turn off the most tired ad, or broaden the audience. Keep a pipeline of 2–3 new creatives ready every couple of weeks.

6. Scale carefully

  • Raise budgets by 20–30% every 3–4 days. Big jumps restart the learning phase and often push costs up.
  • Or duplicate a winning ad set into a new campaign with a broader audience.
  • Try Advantage+ placements, audience and creative once you have conversion data — and for e-commerce, an Advantage+ sales campaign.

Automated rules can help: for example "turn off an ad set if cost per result is above ₹400 in the last 7 days", or "notify me if frequency is above 4".

7. The monthly report

A good monthly report fits on one or two pages:

  • Summary: spend, reach, impressions, results, cost per result, CTR, frequency.
  • This month vs last month with the change in %.
  • Best creatives with their results — clients love seeing which ad worked.
  • Who and where: best age groups, placements and cities.
  • What we changed this month, results of those changes, and the plan for next month.

Build it in Looker Studio or as a simple PDF. Always write the "why" in plain words — numbers without explanation don't build trust.

A weekly optimisation routine

  1. Monday: compare last 7 days with the 7 days before for each ad set — results, cost per result, CTR, CPM, frequency.
  2. Turn off ads that spent at least 2–3× your target cost per lead without a result.
  3. Add 1–2 new creatives where frequency is high or CTR is falling.
  4. Raise the budget by 20–30% on ad sets below your target cost per result.
  5. Check breakdowns once a week, not every day — small numbers jump around.
  6. Write one line in a change log: what you changed and why. It makes the monthly report easy.

Avoid the most common mistake: changing everything at once every day. Meta needs stable settings to learn, and you need stable settings to know what worked.

Practise it

In the Edukatz Meta account, use Plan & optimise: the Budget planner (Day 10), A/B tests (Day 10), Optimise (Day 11) to spot fatigue and scale winners, then Reports (Day 12) to write notes and download your monthly PDF.

Frequently asked questions

How much should I increase the budget when scaling?

About 20–30% every 3–4 days. Bigger jumps can restart the learning phase.

What frequency is too high?

It depends on the audience, but above about 3 in a week with a falling CTR usually means ad fatigue.

How long should an A/B test run?

At least 7 days, with similar budgets and roughly 20 or more results in total.

Edukatz Meta Team

Practical Meta Ads guides by Edukatz trainers — written for beginners and tested in our training accounts.