Five years ago, $200/month of ad spend got you a few hundred clicks and a headache. In 2026, with Meta Advantage+ and Google Performance Max maturing, that same budget can produce meaningful results — if you stop trying to outsmart the algorithm and start feeding it what it needs.

This is the playbook we hand to small business owners running ads on a tight budget. It is opinionated, ignores most of the 'optimisation' tactics that worked in 2020, and focuses on the three things that actually still move the needle in 2026.

The $200 reality

Let us be honest about what $200/month actually buys. On Meta, that is roughly 50-150 link clicks depending on industry. On Google Search, it is 30-80 clicks. On Performance Max, it is harder to predict but lands in a similar range.

That is not enough to optimise tactically. It is enough to test a clear offer, learn whether your audience exists on that platform, and decide if a larger budget would actually pay back. The trap is treating $200/month like $20,000/month — micromanaging audiences, running 15 ad sets, switching strategies weekly. Do that and you will burn $200 for zero learning.

The two tools that matter

For 90% of small businesses, you only need two platforms.

Meta Advantage+ Shopping (or Lead) Campaigns

Advantage+ is Meta's AI-first campaign type. You give it a goal (sales or leads), one creative pool, a daily budget, and minimal targeting input. The algorithm does the rest — and on small budgets it is reliably better than manual targeting now.

Google Performance Max

Performance Max is Google's equivalent. One campaign serves across Search, Display, YouTube, Gmail, and Maps simultaneously. On small budgets the diversity of placements helps; on large budgets people start wanting more control, but you are not at large budgets.

Step 1 — Feed the algorithm 50 conversion events before optimising

The single biggest mistake we see on small budgets: judging campaigns before the algorithm has enough data. Meta Advantage+ needs roughly 50 conversion events to start optimising effectively. Until you hit that number, results look random.

If your conversion is 'purchase' and your average order is $40, $200/month at a 10% conversion rate gives you maybe 4-6 purchases. That is nowhere near 50. So either: (a) switch your conversion event to something cheaper higher in the funnel — 'add to cart' or 'lead form submitted' — until you accumulate 50, or (b) commit to 3-4 months at $200/month before you judge results.

Step 2 — Broad audiences, narrow creatives

The old strategy was the opposite: narrow audiences (lookalike of 1% buyers, in this zipcode, age 28-37, interested in yoga AND nutrition), broad creative (one 'good enough' ad). That world is over.

In 2026 the winning strategy is broad audience (or no audience at all — let the AI find them) and 4-6 distinct creatives that the algorithm can A/B internally. Each creative should hit one specific angle: price, urgency, social proof, founder story, before/after, problem awareness. Let the algorithm pick the angle that works for each viewer.

Step 3 — The 7-day data minimum

Never make optimisation decisions on less than 7 days of data on small budgets. Daily numbers swing wildly when the budget is small — Tuesday might look like a disaster and Thursday like a winner for reasons unrelated to your campaign.

Set up a Friday review slot. Look at the rolling 7-day cost-per-acquisition. If it is dropping week over week, do nothing — the algorithm is learning. If it is flat or rising after 4 weeks, change ONE thing: the creative pool, the offer, or the landing page. Never more than one.

What to ignore

  • Detailed audience targeting on Meta. The algorithm is better than you at finding buyers.
  • Manual bidding strategies on Google. Use Maximise Conversions until you have 100+ conversions/month.
  • Day-parting and dayparts. Meaningless on $200/month — you do not have the data.
  • Negative keyword tinkering on Performance Max. The platform handles it; obsessing wastes hours.

When to scale up

If your blended cost-per-acquisition lands below 30% of your average order value for two consecutive months at $200/month, increase to $400/month. Then $800. Then $1,500. Increase by no more than 50% in a single step — Meta's learning resets if you scale too fast.

If after 3 months at $200/month your cost-per-acquisition is still above your AOV, the issue is not your targeting. It is your offer, your landing page, or the platform fit. No amount of AI optimisation rescues a weak offer.

The bottom line

$200/month on ads is a learning budget, not a growth budget. Treat it that way. Pick one platform (Meta if you sell consumer products, Google if people search for what you offer), commit for 90 days, feed the algorithm 50 conversions, and decide based on rolling 7-day data — not Tuesday's panic.

If after 90 days the numbers say scale, scale. If they say stop, stop. Either answer is worth the $600 you spent learning.