A 4-step framework for proving impact

With AI-driven search and different media channels reshaping the way people discover brands, the “set it and forget it” approach to measuring marketing is officially dead.

Measuring impact is not a static assessment of dashboard data. Used strategically, measurement is a virtuous cycle in which data informs ad platform settings and those settings, in turn, produce better data (and business results).

Here’s how to build a balancing flywheel that keeps your growth running smoothly.

4-step measurement cycle

Consider Bay Area SaaS company PowerLoop, which sells an AI-powered analytics platform. They are investing heavily in Google Search, LinkedIn, and other emerging AI publication funding.

Their problem? Google ads report good ROAS, but their internal CRM shows a significant number of leads and opportunities that cannot be directly attributed to any ad campaign, making it difficult to prove the real impact of marketing across the board.

1. ROAS platform

This is your reality inside the engine. Whether it’s Google Ads or Meta, the ROAS platform uses pixel data and the conversion API to tell you what the platform thinks happened. This may go without saying, but platforms are not in the habit of underestimating their impact.

Eligible: Use this to optimize in real time.

Limitation: These signals feed into your tCPA (targeted cost per acquisition) or tROAS (targeted return on ad spend) bidding strategies. It’s a quick answer you have, but it’s rarely the full truth. This leads us to…

What it looks like in practice (example): The Google Ads PowerLoop account is configured with the tCPA bid strategy “Free trial registration.”

Google Ads reports a healthy $50 CPA, within their target. LinkedIn also shows strong engagement and click rates. This looks good on paper, but unspecified lead is a worrisome concern.

Dig deep: How to avoid marketing model mistakes that hurt results

2. Reverse ROAS

The field data is optimistic. Your bank account is real.

Reverse ROAS, from your CRM of choice (Salesforce, Shopify, HubSpot, etc.), connects your ad spend to your real CRM or internal database. It will probably require some data engineering work to properly map back-end performance versus ad platform waste, but the effort is well worth it.

Eligible: Weed out the “noise” (chargebacks, fake leads, or credit card denials), and evaluate marketing effectiveness based on your first-party data.

Benefit: You can use reverse ROAS to verify your account structure. If the platform says the campaign is winning but the back end is showing low quality leads, it’s time to reorganize your strategy or creative.

What it looks like in practice (example): When PowerLoop connects their ad revenue to Salesforce, they find that “Free trial signups” from Google ads may be incomplete profiles or from IP addresses outside of their target market and never convert to qualified sales opportunities.

LinkedIn, while showing engagement, has a lower than expected conversion rate. This insight leads them to optimize their Google Ads target audience and adjust LinkedIn campaign objectives to focus more on high-intent lead forms.

3. Incremental ROAS (ROAS)

This is the word “So what?” metric. iROAS answers the question: How many of these sales would have happened even if we didn’t show the ad? This is where marketing mix modeling (MMM) and growth testing (geo-lift testing or holdout testing) come into play.

Goal: Get real value and “halo effects” on all channels.

Action: MMM insights tell you where to double down and where to pay for customers who would have converted anyway. Use this information to prioritize your next cycle of growth testing.

What it looks like in practice (example): PowerLoop conducts geo-lift testing by pausing Google Ads in selected non-core markets for a few weeks and measuring the difference in subscriptions between dark areas and similar areas where ads are still running. They found that while Google Ads boosted signups, a large portion of those Googled would have signed up organically anyway, through direct traffic or referrals.

On the other hand, their MMM suggests that AI publishing sponsorship, although it does not drive direct “last click” conversions, has a significant impact on brand awareness and overall CPA reduction across all digital channels by driving more organic searches of their brand. This reveals that the sponsorship has a higher iROAS than originally thought.

Here is an example of less important and less valuable channels:

If the trend is increasing, this channel has become less respected, like YouTube and podcasts in this example. The lower the incrementality factor, the more overvalued these channels are, such as paid review sites in this case.

Dig deep: Why reach is the only metric that proves true marketing impact

4. Minimum ROAS (mROAS)

The final frontier is understanding where to spend the next dollar. Each channel eventually reaches a plateau where efficiency is saturated. This truism is called the law of diminishing returns. Understanding when you get that sign is key to creating an effective budget.

Goal: Measure “room to grow” before hitting the performance ceiling.

Benefit: By monitoring mROAS, you know when to pull back from a full channel and reallocate that budget to emerging areas.

What it looks like in practice (example): PowerLoop analysis shows that after spending $100,000/month on Google Ads, an additional $10,000 yields a minimal return of $0.80 for every dollar spent – ​​meaning they break even on additional spend.

However, with their AI publication funding, each additional dollar spent still returns $2.50 in incremental value, showing significant room for growth. They decided to reallocate 15% of their Google ad budget to expand their sponsorship program.

Why does the cycle never end

Marketing measurement is a work in progress because the landscape is constantly changing. Today, you can improve your Google Search strategy. Tomorrow, you’re thinking about how to measure the impact of what’s being said on a ChatGPT or Perplexity response.

The thoughtful PowerLoop team understands this. They are constantly testing new AI-driven channels and planning how to integrate them into their rating cycle. They know that what worked last quarter may not work this quarter and that relying solely on field data is a waste of money.

The goal is not to find a “perfect” number that stays fixed. The goal is to use this cycle to stay agile. If your iROAS shows that the channel is growing more than you thought, you push your tROAS target on the platform (Step 1) more aggressively. When mROAS shows you’re reaching a high point, you start exploring new, unproven channels to find a different audience.

Dive deep: Breaking down data silos: How integrated analytics can deliver marketing impact