Is AI Ad Targeting Better Than Manual Methods? What E commerce Brands Actually See

An honest breakdown of where AI wins, where it doesn’t, and what the data says in 2025

This question comes up constantly in DTC and e commerce circles. Paid media teams are being told to ‘trust the algorithm’.

especially with emerging AI communication ecosystems while manually managed campaigns are being phased out.

The honest answer is nuanced and it depends on what stage your brand is at, what data you have, and what you’re actually trying to optimise for.

What AI Ad Targeting Actually Means

‘AI targeting’ is used loosely to describe several different things:

• Automated bidding (Google’s Smart Bidding, Meta’s Advantage+) the algorithm decides when and how much to bid based on conversion signals
• Audience expansion the algorithm finds users similar to your converters without you defining lookalikes manually
• Creative optimisation AI tests variations of your ad creative and serves the best performers automatically
• Predictive segmentation AI identifies which users in your existing customer base are most likely to buy again (useful for re engagement campaigns)

These are related but distinct capabilities. Whether AI outperforms manuals depends entirely on which of these you’re comparing.

Where AI Targeting Demonstrably Wins

Bidding and budget optimisation

Google’s Smart Bidding and Meta’s AI bidding systems process millions of real time signals device, time, query intent, user history, location that no human can manually process at scale. For campaigns with sufficient conversion data (30+ conversions per month is typically the threshold), automated bidding consistently outperforms manual CPC management.

The catch: these systems need data to learn. New accounts, small budgets, and low conversion products starve the algorithm. In those cases, manual bidding often outperforms AI in the early weeks.

Audience discovery

Meta’s Advantage+ audience tool and Google’s Broad Match with Smart Bidding have shown genuine improvements in finding converting users that human built lookalikes miss. Multiple DTC brands have reported lower CPAs after removing manual audience restrictions and letting the algorithm expand.

The underlying logic is sound: human marketers build audiences based on known segments. AI finds converting users based on behaviour patterns that are invisible to humans. These aren’t the same thing, and both have value.

Creative testing at scale

Dynamic creative optimisation (DCO) where AI mixes and matches headlines, images, and CTAs to find top performers consistently outperforms manual A/B testing because it can run more variations in parallel and optimise faster. Brands running 10+ creative variants benefit most from AI managed testing.

Where Manual Still Has an Edge

Brand control and messaging alignment

AI Voice Agent optimises for conversions, not brand coherence. Left unchecked, automated creative systems will find the message that converts which may not be the message you want associated with your brand. Manual oversight remains essential for keeping ad creative on brand, especially for DTC brands where brand voice is a competitive asset.

New product launches and low data environments

AI systems need historical conversion data to optimise effectively. A new product, a new market, or a heavily restricted audience gives the algorithm too little signal. Manual targeting testing defined segments deliberately is often more efficient in the early phase.

High AOV, low volume purchases

If your product sells 20 units a month, your conversion data is thin. AI bidding systems may behave unpredictably or over optimise on proxy signals (clicks, add to carts) rather than actual purchases.

Manual campaign management with tightly controlled audiences often produces more reliable results in these cases.

The Real Question: Paid Acquisition vs Owned Channel ROI

Here’s the thing most ‘AI vs manual targeting’ discussions miss: ad targeting AI or manual is about acquiring new customers. But the brands with the highest growth are winning on owned channels, not paid. AI tools that operate on your existing customer base often deliver better ROI than ad campaigns at any budget level.

Two examples of this in practice:

• AI personalisation engines (Rebuy, LimeSpot, Nosto) increase revenue from existing visitors by serving the right product at the right moment no ad spend required
• AI phone platforms (AI voice agents) recover revenue from high intent customers who already visited your store and abandoned targeting people who were close to buying, not cold audiences

For many Shopify brands, the fastest path to revenue isn’t better ad targeting, it’s better monetisation of the traffic they already have.

Consio and the ‘Zero Ad Spend’ Recovery Problem

Cart abandonment represents roughly 70% of all e commerce sessions. Those are visitors your paid ads already acquired. When they leave without buying, you’ve already paid for the acquisition and most of that revenue is walked out the door.

Consio’s AI phone platform works on this ‘already acquired but not converted’ segment  calling Shopify abandoners within minutes, handling objections in real time, and recovering orders that email and SMS alone can’t reach.

Medterra ($25K in 21 days)  sees more real customer case studies and AndorWillow ($1M annualised) have turned this into a material revenue channel.

From a pure ROI standpoint, recovering 40% of high value abandoners (EVOLV’s documented rate) often delivers better returns than the same budget spent on incremental ad targeting improvements.

Practical Guidance: When to Lean on AI, When to Stay Manual

Scenario	Recommended approach
Campaign with 50+ monthly conversions	Use AI bidding (Smart Bidding / Advantage+)
New product, < 20 conversions/month	Start manual, build conversion data first
Creative testing across 10+ variants	Let AI DCO optimise automatically
Brand campaign requiring tone control	Manual oversight of AI generated creative
Re engaging existing customer base	AI segmentation + owned channel outreach
Recovering abandoned high AOV carts	AI phone recovery (Consio AI) > paid retargeting

If you’re spending on paid acquisition but not recovering abandoned carts, you’re paying twice for the same customer. request a demo