Paid Media

Bid Strategies in Google Ads: Manual, Maximize Conversions, or tCPA — Which to Use and When

By Haris Ahmed Siddiqui · · 12 min read

If you are spending real money on Google Ads in Pakistan and your cost per lead keeps creeping up, the bid strategy you picked is usually part of the problem. This guide is for business owners and in-house marketers who manage their own campaigns (or want to sanity-check the agency that does) — and it gives you a clear decision tree for choosing between manual CPC, Maximize Conversions, and Target CPA based on the one thing most people ignore: how much conversion data your account actually has.

We have run paid campaigns for Pakistani clients since 2009, and the single most expensive mistake we see is people copying a bid strategy off a YouTube video meant for a US e-commerce store with 500 sales a month. Your account is not that account. Let’s fix the choice properly.

Why google ads bid strategies decide your results more than your budget

People obsess over budget and keywords. But the bid strategy is what tells Google’s system how to spend every rupee of that budget in every single auction. Pick the wrong one and you can double your spend without doubling your leads. Pick the right one for your data maturity and the same budget quietly gets more efficient over time.

Here is the mental model that matters. There are two families of google ads bid strategies:

  • Manual bidding — you (or your account manager) set the maximum cost-per-click. You are in control. The machine does not optimise toward conversions automatically.
  • Smart bidding — Google’s machine learning sets bids in real time for each auction, using signals you cannot see manually (device, time of day, browsing history, location, and hundreds more). This family includes Maximize Conversions, Target CPA, and Target ROAS.

Smart bidding is genuinely powerful, but it is not magic. It runs on your conversion data. No data, no magic — just expensive guessing. That is the whole game, and it is why the rest of this article is organised around conversion volume rather than around the strategies themselves.

The honest decision tree: pick by conversion volume, not by hype

Forget “best practices” for a second. Answer one question: how many conversions does your account record per month? A conversion here means a real lead — a form fill, a WhatsApp click, a JazzCash/Easypaisa checkout, a verified phone call. Not a pageview.

0–15 conversions per month: start with Manual CPC (or Maximize Conversions, carefully)

If you are launching a new campaign with no conversion history, smart bidding has nothing to learn from. We usually start here with manual CPC for the first few weeks. You set sensible click bids, watch which search terms convert, and prune the junk. This is unglamorous but it keeps you in control while data accumulates.

If you want to lean toward automation early, Maximize Conversions is the next gentlest step — it does not require a CPA target, so it works on thin data. But watch your budget like a hawk, because Maximize Conversions will spend your entire daily budget by design. Set the daily cap at a number you can afford to lose while it learns.

15–30 conversions per month: Maximize Conversions, then graduate

This is the awkward middle. You have enough data for the system to find patterns, but not enough for tight targets. Run Maximize Conversions and let it gather at least 30 conversions in a rolling 30-day window before you even think about adding a target. Resist the urge to bolt on a CPA target the moment leads come in — that is the trap we cover below.

30+ conversions per month: Target CPA is now on the table

Google’s own guidance — and our experience — says target CPA needs roughly 30 conversions in the past 30 days to work reliably. Below that, the target becomes a leash that strangles delivery. At or above it, tCPA lets you say “get me leads at PKR 800 each” and the system optimises toward that. This is where smart bidding starts earning its keep for most lead-gen businesses in Pakistan.

50+ conversions per month with revenue data: consider Target ROAS

If you are an e-commerce store passing back purchase value (so Google knows order #1 was PKR 2,400 and order #2 was PKR 18,000), target roas bidding becomes viable. ROAS bidding optimises for revenue, not just lead count — it will happily spend more to win a high-value cart and pull back on cheap ones. But it needs more data than tCPA and clean value tracking. Get the tracking wrong and ROAS bidding optimises toward fiction.

The biggest trap: switching strategies too early

This deserves its own section because it is the mistake that burns the most money. Every time you change bid strategy — or significantly change the CPA/ROAS target — Google enters a learning period, typically 7 days but sometimes up to two weeks. During learning, performance is volatile and often worse. The algorithm is recalibrating.

Here is what impatient advertisers do. They launch tCPA on Monday. By Thursday cost per lead looks ugly (because it is still learning), they panic, switch back to Manual, then to Maximize Conversions next week, then back to tCPA with a lower target. They never let any strategy stabilise. The account spends its whole life in learning mode, which is the most expensive mode there is.

Our rule, learned the hard way:

  • Give any new bid strategy a minimum of 2 weeks plus 30 conversions before you judge it. Whichever comes later.
  • Change one thing at a time. Do not change strategy, budget, and targets in the same week — you will never know what worked.
  • When adjusting a tCPA or ROAS target, move it in steps of 15–20%, not 50%. Big target jumps trigger fresh learning and tank delivery.
  • Do not switch strategies during a sale, Ramadan demand spike, or any abnormal traffic period. Wait for normal conditions.

If you only remember one thing from this article: patience is a bid strategy. The data has to mature, and switching too early is how you guarantee it never does.

How each strategy actually behaves day to day

Manual CPC (and Enhanced CPC)

Full control, full responsibility. Good for tiny budgets, brand-new accounts, and situations where you genuinely understand your keywords better than the machine does (very rare once you have data). Enhanced CPC is a half-step — you set manual bids and let Google nudge them up or down for likely conversions. We treat ECPC as training wheels, not a destination.

Maximize Conversions

Get the most conversions possible for your daily budget. No target, so it works on thin data. The catch: it will always spend the full budget, and without a target it does not care whether a lead cost PKR 300 or PKR 3,000 — only that it got one more. Excellent for the early-data phase; risky if you set the budget too high before you trust it.

Target CPA

The workhorse for Pakistani lead-gen — services, real estate, education, clinics. You name a cost per lead and the system chases it. Needs that 30-conversions floor. If your target is too aggressive (far below your historical actual CPA), delivery collapses and your ads stop showing. Set the first target near your current real CPA, then tighten gradually.

Target ROAS

For e-commerce with value tracking. Optimises return on ad spend. The most data-hungry and the least forgiving of broken tracking. Brilliant when your product margins vary and you want the system to prioritise profitable orders — pointless if every order has the same value or your pixel is misfiring.

The thing that makes all of this work: conversion tracking

None of these strategies function without accurate conversion tracking. Smart bidding optimises toward whatever you tell it is a conversion. If you tag every “Contact Us” pageview as a conversion, the algorithm will dutifully buy you thousands of useless pageviews and report a fantastic cost per “conversion” while your actual sales stay flat.

Before you touch bid strategy, get these right:

  • Track real outcomes — verified leads, qualified form fills, completed checkouts, WhatsApp click-to-chats that actually start conversations.
  • De-duplicate. One enquiry should be one conversion, not three because the user clicked WhatsApp twice.
  • For e-commerce, pass actual order values back so ROAS bidding has something true to optimise on.
  • Watch for the mid-range Android reality in Pakistan — slow connections and back-button behaviour can break tracking tags. Test on a real cheap phone, not just your laptop.

Conversion tracking and a landing page that actually converts go hand in hand. If your traffic is fine but leads are thin, the problem may be the page, not the bid strategy — our web design and landing page team handles that side, and our PPC management service wires the tracking before we ever change a bid setting.

A practical migration path we use for clients

Here is the sequence we run for a typical Pakistani lead-gen account starting from scratch:

  1. Weeks 1–3: Manual CPC or Maximize Conversions with a tight daily budget. Build keyword and search-term data. Kill irrelevant queries with negatives.
  2. Weeks 3–6: Once 30 conversions are banked in a rolling month, switch to Target CPA with the target set at — or slightly above — current actual cost per lead. Then leave it alone for two weeks.
  3. Weeks 6–10: Tighten the tCPA target in 15% steps as performance allows. One adjustment at a time, two weeks of breathing room between each.
  4. Ongoing: Only consider Target ROAS if the account becomes value-based e-commerce with clean revenue tracking and 50+ monthly conversions.

This is deliberately slow. Slow is cheaper than churn. Every shortcut here costs you a fresh learning period.

Frequently Asked Questions

I have a brand-new Google Ads account. Should I use smart bidding right away?

No. With zero conversion history, smart bidding has nothing to learn from and will spend inefficiently. Start with Manual CPC or Maximize Conversions on a small daily budget, gather data for a few weeks, and graduate to Target CPA once you have around 30 conversions in 30 days. Jumping straight to tCPA on a cold account is one of the most common ways to waste your first month’s budget.

How many conversions do I really need for Target CPA to work?

Roughly 30 conversions in the trailing 30 days as a practical floor — Google’s own threshold. Below that, the target restricts delivery and your ads may stop showing. If you are nowhere near 30, stay on Maximize Conversions until the volume builds. Forcing tCPA on thin data is the fastest way to choke a campaign.

My cost per lead got worse after switching strategies. Did I pick the wrong one?

Probably not — you likely judged it during the learning period. Every strategy change triggers 7–14 days of volatility while the algorithm recalibrates. Give any new strategy at least two weeks and 30 conversions before deciding. If you keep switching back and forth, the account never leaves learning mode, and that is what actually drives your cost up.

Is Maximize Conversions going to blow my whole budget?

Yes — by design it spends your full daily budget chasing the most conversions it can get. That is fine if your daily cap is set at a number you are comfortable spending. The risk is setting the budget too high before you trust the campaign. Cap it conservatively while it learns, then raise it once the cost per lead is acceptable.

What about Target ROAS — is it better than Target CPA?

Neither is “better”; they solve different problems. Target CPA optimises for the number of leads at a cost you set, which suits most service and lead-gen businesses. Target ROAS optimises for revenue and only makes sense for e-commerce that passes accurate order values back to Google. If every order is roughly the same value, tCPA is simpler and just as effective.

Can I just leave it on Manual CPC forever and avoid all this?

You can, but once you have real conversion data you are usually leaving money on the table. Smart bidding reads signals — device, time, location, intent — that you cannot adjust for manually in every auction. Manual CPC is the right starting point and a fine fallback for tiny budgets, but for most accounts with 30+ monthly conversions, a well-set Target CPA outperforms manual management.

Get your bid strategy chosen the right way

Choosing between manual CPC, Maximize Conversions, Target CPA, and Target ROAS is not about which one is trendiest — it is about matching the strategy to your conversion volume and your tracking maturity, then having the discipline not to switch too early. Most accounts we audit are either using a strategy their data cannot support, or thrashing between strategies and living permanently in learning mode.

If you want a second opinion, One Source Soft offers a free audit of your existing account — we will tell you honestly whether your current bid strategy fits your data, where your tracking is leaking, and what we would change first. Our clients leave the kind of reviews you can read publicly on Google, and we would rather earn yours by saving you money than by making you a slide deck. Talk to our team through our PPC management service or reach out directly via our contact page to book the consultation.

Related reading

Keep reading