"Advice for tCPA adjustments" is a perennial request in practitioner forums, and for good reason: Target CPA is the default bid strategy for most B2B SaaS Search campaigns, yet the act of changing the target is where accounts most often self-inflict damage. Move it too far, too fast, or too often and you can crater conversion volume or send CPA the wrong way — then blame the strategy for a problem the adjustment created. The mechanics are not complicated, but they are unforgiving of impatience.
This is a tactical guide to the how: how big a step to take, how long to wait, what conversion volume you need before touching the target at all, and how to read the aftermath so your next move is informed rather than reflexive. It assumes you have already chosen Target CPA deliberately; if you are still deciding between bid strategies, start with our overview of Google Ads bidding strategies for B2B SaaS and come back when the target itself is what you need to tune. A recent Google change, covered at the end, makes getting this right more important than it was even a few months ago.
Volume before targets: the 30-conversion floor
Before you touch the target, check whether the campaign has the conversion volume to respond to one predictably. The widely cited threshold for stable Target CPA performance is 30 or more conversions per month. Below that, the algorithm is working from too little signal, its results swing from week to week, and any target change you make is being interpreted by a bidder that cannot reliably distinguish a good auction from a bad one. Tightening the target on a data-starved campaign does not make it efficient; it makes it erratic.
For B2B SaaS this floor is harder to clear than it looks, because the conversions that matter — SQLs, opportunities, closed-won — are rarer and slower than form fills. The right response when volume is thin is not to keep adjusting the target but to fix the signal: import earlier, higher-in- the-funnel conversions like booked demos and SQLs through your offline conversion stack, or consolidate fragmented campaigns so the conversion data concentrates instead of scattering across too many thin ad groups. Get the campaign above the volume floor first; only then does target tuning become a productive lever rather than a source of noise.
Setting the initial target: anchor to reality
The first target should be anchored to what the account already achieves, not to what you wish it achieved. The standard practice is to set the initial Target CPA at your actual recent CPA — if the campaign has been converting at roughly $180 over the last 30 days, start the target near $180. Going 10 to 20 percent above your historical average gives the algorithm room to learn without immediately constraining it, which matters because a target set aggressively low on day one often prevents the bidder from ever gathering the data it needs to improve.
The temptation to set an ambitious target from the outset — "we need leads at $120, so let's just set $120" — is the most common opening mistake. If the account is not already converting near that number, the bidder responds by pulling out of most auctions in a doomed attempt to hit an unrealistic cost, and volume collapses before any learning happens. Anchor to reality, let the strategy stabilize, and improve the target gradually from there. The target is a steering input, not a wish; the algorithm treats it as a literal instruction about which auctions are worth entering.
Step size: move in 10–20% increments
Once a campaign has exited learning and stabilized, adjust the target in small steps — 10 to 20 percent per change, with 10 to 15 percent a safe default when you want to protect efficiency. Small increments let you probe the trade-off between cost and volume without shocking the system. Each move shifts which auctions the bidder is willing to enter; a modest change nudges that boundary, while a large one yanks it, often triggering an abrupt volume swing and a fresh bout of instability as the algorithm re-learns.
Direction matters as much as size, and it forces an honest choice about the goal. Lowering the target chases efficiency at the expense of volume, because the bidder stops entering the more expensive auctions — some of which convert. Raising it buys volume at higher cost. You cannot get cheaper conversions and more of them from the same move; that is the core tension of Target CPA. Decide which you are optimizing for before you touch the number, and if the answer is "more pipeline," be prepared to raise the target rather than reflexively cut it. For the deeper strategic framing of that trade-off, see manual bidding vs. automated strategies.
Cadence: wait for conversions, not calendar days
The single most violated rule of Target CPA management is cadence. Do not adjust more often than roughly every two weeks, and gate each change on conversions collected, not days elapsed. When you change a target the bidder reacts immediately, but the conversions that tell you whether the change worked take one or two conversion cycles to materialize. Adjust again before those conversions arrive and you are steering on noise, overwriting a change whose result you have not yet seen.
B2B SaaS stretches this waiting period further than most verticals, because the conversions you optimize toward often arrive weeks after the click and, when imported offline, later still. A campaign optimizing to SQLs on a 60-day cycle simply cannot be evaluated on a one-week look. This is why attribution setup and target cadence are linked: if your conversion window is too short, conversions never fully land and you will misread every adjustment as underperforming. Set the window to match the cycle, then wait for a genuine sample — commonly 20 to 30 conversions at the new target — before you judge the move and plan the next one.
Reading the aftermath before the next move
After a change and an adequate waiting period, read three things together before deciding anything: conversion volume, achieved CPA, and whether the campaign spent to its budget. Interpreting CPA alone is how teams talk themselves into bad moves. A lower achieved CPA looks like success, but if volume fell more than proportionally you may have traded away profitable conversions to make an average look better. Conversely, a higher CPA alongside a jump in qualified pipeline can be exactly what you wanted, if the incremental conversions clear your LTV:CAC threshold.
Budget interacts with all of this, and it is about to interact more. If a campaign is budget-limited, its behavior against the target is not what the target alone would predict — a nuance that becomes central under the August 2026 change below. Whenever possible, evaluate a target change on a campaign that is not pinned against its daily budget, so you are measuring the target's effect and not the budget cap's. When budget and target are both binding, resolve the budget question first; there is little point tuning a target the algorithm cannot fully act on because it runs out of money by midday.
The August 2026 change and why hygiene now matters more
Google is rolling out a Smart Bidding change from August 17, 2026 that makes target hygiene more consequential than it used to be. Under the update, budget-limited Target CPA and Target ROAS campaigns will optimize more closely toward the target you actually set, rather than overdelivering below it as many budget-constrained campaigns previously did. In plain terms, a loosely set target you have been ignoring may now be taken more literally by the system, which can change delivery on campaigns you had not planned to touch.
The practical response is to audit your targets against recent delivery before the change fully lands, and to use Google's Bid Target Adjustment Tool to choose deliberately: keep the current target and accept the drift, lower it to match recent performance, or set a custom figure between the two. Our walkthrough of the August 2026 bidding target optimization covers the mechanics in full. The through-line is that everything in this guide — anchor to real CPA, move in small steps, wait for conversion volume, read the full picture — is precisely the discipline the new behavior rewards. Advertisers who set targets carefully were always better off; after August, the ones who set them carelessly will simply feel it faster.