Smart Bidding for Small Shopify Stores: tCPA vs tROAS vs Maximize
How small Shopify stores should choose between Target CPA, Target ROAS and Maximize conversion value bidding, with break-even math.
Written by Mantas Jurgutis — Founder, Adsify — builds the Google & Meta automation merchants use daily
Editorially reviewed by Adsify Editorial on April 10, 2026 — Reviewed against Shopify, Google Ads and Meta official documentation.
Why bid strategy choice matters more for small accounts
Smart Bidding strategies all rely on machine learning models trained on your account's conversion history, and small Shopify stores with limited monthly conversions give those models less data to learn from than large accounts, which makes the choice of strategy — and the timing of when you introduce a target — more consequential. Picking Target ROAS with an unrealistic target on a store doing 10 conversions a month typically throttles spend severely, while Maximize conversion value with no target lets the algorithm explore more freely while data accumulates.
Maximize conversion value: the right starting point
Maximize conversion value, without a target ROAS, tells Google Ads to spend your full daily budget while maximizing total conversion value, without constraining efficiency to a specific ratio. This is the recommended starting strategy for stores with fewer than 30 conversions in the trailing 30 days, since it lets the algorithm gather data across a wider range of auctions rather than being restricted early by a target that has no historical basis. Expect ROAS to fluctuate more in this phase — that volatility is largely the cost of the learning process, not a sign something is broken.
Target ROAS: when you have enough history to set one
Once you have roughly 30+ conversions in the last 30 days (Google's general guidance for reliable Target ROAS performance, though it can work with less), you can layer in a tROAS target based on your account's recent actual performance rather than a number you'd like to hit. If Maximize conversion value has been running at a natural 380% ROAS, setting an initial tROAS target around 350-380% (roughly matching or slightly below current performance) avoids an abrupt spend cut, then you can raise the target gradually — Google recommends adjusting by no more than 10-15% at a time to avoid destabilizing the bid model.
Target CPA: when it fits better than tROAS
Target CPA optimizes toward a cost-per-conversion goal rather than a value-based ratio, which suits stores where most orders have similar value (so optimizing for volume of conversions roughly equals optimizing for revenue) or where you're tracking a non-purchase conversion like email signups or a lead form alongside purchases. For catalogs with wide price variance — a $15 accessory and a $300 flagship item in the same campaign — tROAS or Maximize conversion value generally serves margin goals better than tCPA, since tCPA doesn't account for the different profit each order represents.
Worked example: setting a break-even tROAS
Assume AOV of $55, cost of goods and fulfillment at 40% of revenue ($22), leaving $33 gross margin per order. If you want ad spend capped at 50% of gross margin to leave room for other overhead, target ad spend per order is $16.50, giving a target ROAS of $55 / $16.50 ≈ 333%. Your true break-even ROAS (where ad spend consumes 100% of gross margin) is $55 / $33 ≈ 167% — useful to know as a hard floor, since any campaign reporting above 167% ROAS is still contributing positive gross margin even if below your 333% target.
Worked example: comparing tCPA and tROAS on the same account
Take a store with two products: a $30 item at 60% margin ($18 profit) and a $120 item at 30% margin ($36 profit). A tCPA of $12 applied uniformly would look efficient on the $30 item (a $12 CPA against $18 profit is comfortable) but risks the $120 item being under-bid relative to its higher absolute profit, since tCPA doesn't distinguish the two. A tROAS target instead scales the acceptable spend with each item's actual price, generally producing better-aligned bids across a catalog with wide price variance like this.
How often to adjust targets
Wait at least 2 weeks or one full conversion cycle (whichever is longer) between target adjustments, and change the target by no more than 10-15% per adjustment. Google's Smart Bidding documentation notes that each meaningful bid strategy or target change can trigger a new learning period, during which performance may be less stable — frequent small tweaks made out of impatience are one of the most common self-inflicted performance problems in small accounts.
Portfolio bid strategies versus campaign-level targets
Small stores with only one or two active campaigns rarely need a portfolio bid strategy (which manages a shared target across multiple campaigns) — set targets at the individual campaign level for simplicity. Portfolio strategies become more useful once you're running 3+ campaigns you want optimized toward a single blended account-wide ROAS or CPA goal, typically past the point most stores using a single PMax campaign need to worry about.
Seasonal adjustments without breaking the model
Ahead of a known high-volume period like a site-wide sale, use seasonality adjustments (Tools and Settings > Bid strategies > Seasonality adjustments) rather than directly changing your tROAS or tCPA target — this tells Smart Bidding to expect a temporary conversion rate change for a specific date range without altering your long-term target, and it reverts automatically afterward. Manually loosening a target for a sale and forgetting to tighten it back afterward is a common cause of margin erosion that persists well past the promotional period.
Recognizing when Smart Bidding is starved of data
Signs your account doesn't yet have enough conversion volume for a stable target strategy include: ROAS swinging by more than 50% week to week with no external cause, budget consistently under-spent despite a reasonable daily cap, or the bid strategy status page showing 'Learning' for more than 2-3 weeks continuously. In these cases, drop back to Maximize conversion value without a target, or Maximize conversions without a tCPA, until conversion volume grows.
The role of value rules for a more nuanced Maximize strategy
Value rules (Tools and Settings > Value rules) let you adjust the value Smart Bidding attributes to conversions based on factors like new versus returning customer, device, or location, without switching away from Maximize conversion value. For a Shopify store where new customer acquisition is worth more strategically than a repeat purchase (due to lifetime value), a value rule that multiplies new-customer conversion value by 1.3-1.5x can shift bidding to favor acquisition without needing a separate campaign structure.
Combining POAS thinking with Google's native metrics
Google Ads only sees revenue, not your actual profit margin, unless you build margin into the conversion value itself — some merchants upload profit (POAS: profit on ad spend) rather than revenue as the conversion value, achieved by adjusting the value passed at checkout to reflect margin rather than price. This is more advanced and requires careful implementation (typically via a script or app that recalculates margin per order dynamically), but it lets tROAS targets optimize toward actual profitability rather than top-line revenue, which matters more as product margins vary.
How Adsify approaches bid strategy for small stores
Adsify's optimizer tracks profit and POAS explicitly rather than only revenue-based ROAS, and adjusts campaign settings on a 6-hour cycle as new conversion data comes in — useful specifically for smaller stores where a single day's order pattern can otherwise skew a human reviewer's weekly manual check.
A simple decision checklist
Under 30 monthly conversions: use Maximize conversion value, no target. 30+ conversions with wide price variance across SKUs: introduce tROAS based on recent natural performance. 30+ conversions with fairly uniform order values: tCPA is a reasonable simpler alternative. Approaching a known sale period: use seasonality adjustments rather than touching your core target. Above all, resist adjusting more than once every two weeks unless you're reverting an adjustment that clearly wasn't working.