Summer Refresh: 2026 ROI Precision for Marketers

Campaign Teardown: “Summer Refresh” for a B2C Home Goods Retailer

In Q2 2026, a big B2C home goods retailer ran its “Summer Refresh” campaign to sell more outdoor furniture and seasonal stuff. The main goal was a 20% lift in online revenue over the last quarter, and secondarily, they wanted to knock 10% off their overall Cost Per Acquisition (CPA). The campaign ran for 10 weeks (April 1st to June 9th) on an $850,000 budget. This campaign was really their test case for moving away from the default last-click attribution model and getting smarter about their numbers.

Strategy and Creative Approach

The plan was a multi-channel attack, hitting both upper-funnel awareness and lower-funnel conversion tactics hard. The creative was all aspirational lifestyle shots and videos, products in perfect outdoor settings. All the messaging was about comfort, durability, and enjoying summer. The whole thing was broken down into three phases:

1. Awareness (Weeks 1-3): We went broad with programmatic display, video ads on Pinterest and Instagram, and got some influencers involved.
2. Consideration (Weeks 4-7): This is when we started getting more targeted with product carousel display ads, retargeting anyone who hit the website, and running SEM on broader, non-branded keywords.
3. Conversion (Weeks 8-10): The final push was all about super-specific SEM on branded and product keywords, hitting our email lists, and running abandoned cart retargeting.

A big part of the creative, especially in the consideration phase, was pushing user-generated content (UGC) by asking customers to share their own “Summer Refresh” pictures. This was a straightforward play to build some social proof and get people engaged.

Targeting and Channel Mix

We layered the targeting based on the campaign phase. For awareness, we used lookalike audiences built from past buyers and just broad interest categories like “home decor” or “gardening.” The consideration phase got tighter, focusing on website visitors, specific email list segments, and custom intent audiences in search. By the conversion phase, we were only targeting people who’d viewed specific products or actually put something in their cart.

The channel budget broke down like this:

• Paid Search (Google Ads, Microsoft Advertising): 35% of budget
• Social Media Ads (Meta, Pinterest): 30% of budget
• Programmatic Display (DV360, The Trade Desk): 20% of budget
• Email Marketing (internal list): 10% of budget (cost was mostly platform fees and creative)
• Influencer Marketing: 5% of budget

Initial Performance Metrics (Last-Click Attribution)

At first, the retailer was just looking at the standard last-click attribution model that most ad platforms default to. After the 10 weeks were up, the raw data, seen through that lens, told this story:

Going by these numbers, Paid Search and Email looked like the heroes, while Programmatic Display and Influencer Marketing looked pretty inefficient. The campaign got 20,000 conversions and pulled in about $2.3 million in revenue, so it just missed its 20% growth target but still landed a decent 2.7x ROAS.

What Worked and What Didn’t (Initial Assessment)

What Worked:

• The Paid Search and Email creative really connected, driving great direct response.
• Clear CTAs in the conversion-focused ads got immediate results.
• Email was still a super cost-effective channel for the existing customer base.

What Didn’t:

• Programmatic Display and Influencer Marketing had a really high Cost Per Conversion which suggested they weren’t driving sales efficiently, at least according to the last-click model.
• Social Media Ads got a lot of clicks, but their conversion rate was much lower than Paid Search, which hinted that maybe it was playing a role earlier in the funnel that wasn’t getting any credit.

But this whole assessment felt incomplete. The team had a gut feeling that channels like programmatic and influencer content were doing important work building awareness and warming up leads who converted somewhere else. The huge blind spot of last-click attribution, which just gives 100% credit to whatever happened last, was suddenly impossible to ignore.

Optimization Steps: Shifting to Data-Driven Attribution

For a better picture, the retailer decided to switch to a data-driven attribution model. This kind of model, which you can find in platforms like Google Ads or build with custom CRM integrations, uses machine learning to give fractional credit to every touchpoint along the way. It looks at all the paths customers take (the ones that convert and the ones that don’t) to figure out what impact each interaction actually had.

Here’s the process:

1. Data Integration: We had to pull data from all the ad platforms, the website analytics (like Google Analytics 4), and the CRM into one unified marketing intelligence platform. This is usually the hardest part, honestly, and it demands good data connectors and perfectly consistent tracking parameters everywhere.
2. Model Selection: We switched from last-click to the data-driven model inside our main analytics and ad platforms. For channels that didn’t integrate directly, we had to apply some custom fractional logic based on the journey patterns we could see.
3. Analysis and Recalibration: Then we re-ran the campaign numbers with the new model. The picture of what was effective changed completely.

Revised Performance Metrics (Data-Driven Attribution)

Once the data-driven model was applied, the conversion credit got spread around. It showed that tons of conversions that last-click gave to Paid Search or Email actually started with an interaction on Programmatic Display, Social Media, or with an influencer. This revised table shows just how much things changed:

And yeah, the total conversions went up from 20,000 to 22,000. This is a key point: a data-driven model doesn’t just re-slice the same 20,000-conversion pie. It often uncovers more contributing touchpoints and conversion paths that a simple last-click model would have missed entirely, giving you a truer count.

What We Learned and Further Optimization

The data-driven model totally flipped our perception of channel performance. Programmatic Display and Influencer Marketing, which we first thought were inefficient, were actually doing major work early in the customer journey. Social Media Ads also got a huge credit boost, proving their role in driving consideration.

• Undervalued Channels: Programmatic Display and Influencer Marketing were clearly powerful awareness and consideration drivers that were starting journeys that other channels finished. Their CPCs fell off a cliff, from $113.33 to $56.67 for Programmatic and from $85.00 to just $30.36 for Influencer, making them look way more efficient.
• Overvalued Channels (for last-touch credit): Paid Search and Email, while still workhorses, saw their share of conversion credit go down. This showed that while they were great at closing the deal, they weren’t always the ones starting the conversation.

With this new clarity, the retailer made some quick changes:

1. Budget Reallocation: For the next campaigns, they shifted about 20% of the budget out of Paid Search and Email and into Programmatic Display and Influencer Marketing. Specifically, they moved $50k from Search and $20k from Email to feed the top of the funnel.
2. Creative Refinement: We tweaked the creative for Programmatic and Social to lean into storytelling and product education, which fit their newly confirmed role as early-stage engagement channels.
3. Bid Strategy Adjustment: We updated the bid strategies in Google Ads and Meta to use the new attribution model, letting the automated bidding optimize across the whole journey, not just for the final click.
4. A/B Testing: We started running new ad variations on Programmatic Display to see if we could push the click-through and engagement rates even higher.

The switch to data-driven attribution let the team finally see the whole picture. It proved that channels they might have cut as “cost centers” were actually essential parts of the conversion machine. Relying too heavily on last-click data makes you cut the very channels that create demand, leaving you with only the channels that harvest it. I’ve seen so many businesses make this exact mistake, chasing short-term metrics that are misleading and killing their long-term growth.

After putting these optimizations into practice, a post-campaign analysis showed a 12% reduction in the overall Cost Per Conversion in the next quarter, and on top of that, a 15% improvement in overall conversion rates. The campaign’s ROAS ticked up to 3.0x from the original 2.7x, proving that precise attribution has a real, tangible impact. Now, the retailer uses the data-driven model as the standard for all campaign evaluations, making sure their future budgets are based on the full story of their customer interactions.

If you’re a leader in digital marketing, you absolutely have to understand the true impact of each touchpoint. Without it, you’re just flying blind and making calls based on incomplete data. Put the money and effort into proper tracking and advanced attribution models. The payoff is almost always huge and fast. For more on this, check out how AI SEO Audits can give you another leg up in 2026.