Measuring Marketing Success

Marketing has two big jobs, but we're usually judged on only one. Our job in marketing splits into two parts: Building mental availability: making sure people know who we are and remember us when they’re ready to buy. This is often called brand marketing. Activating demand: making sure that people who are ready to buy choose us. This is typically performance or demand marketing. Here’s the challenge — most of our metrics (MQLs, pipeline, revenue) are tied to demand activation. But brand and demand aren’t separate – they work together. Still, they behave differently and aren’t always easy to measure in the same way. Brand is like staying in shape. You go to the gym, eat healthy, and take care of yourself. You don’t always see instant results, but over time, your body gets stronger. → In marketing terms: We want more people to know us, remember us, and think of us when they’re ready to buy. This is a long-term game. Demand activation is like showing up on race day. You’ve trained for months, and now it’s time to perform. If you’re fit, you’ll likely do well. → In marketing terms: When someone’s ready to buy, our goal is to be easy to find and hard to ignore. Most of the time, our execs care about the race day numbers – leads, opps, deals. That’s fair, because those drive revenue. But if we don’t also take care of our brand (our fitness), performance eventually suffers. So what do we do? We need to measure both. Performance marketing already has clear metrics. But brand often feels fuzzy — hard to prove it’s working. That’s why Share of Search (SoS) is useful. It’s a quantifiable way to track how much people are searching for our brand compared to competitors. It acts like a “brand scoreboard”, so we can see how campaigns are moving the needle, even if the revenue impact comes later. So: Use performance metrics for activation (leads, opps, CAC, etc.) Use Share of Search as the north star for brand Run both in parallel, and know that each supports the other Two different motions. Two different metrics. One goal: revenue growth.

Have written about it in the past too, but strength of a brand cannot be measured by vanity metrics like number of award winning commercials, social media engagement/following or what other marketers think about the brand Here are 5 hard measurable business/marketing metrics which will tell you how strong your brand is 1. Price Elasticity of Demand: This is the measurement of change in demand with respect to the Price Elasticity= Percentage Change in Demand/Percentage Change in Price If you have a strong brand, you will have a lower price elasticity. Ideally as brand strength grows, the price elasticity should keep reducing 2. Contribution of Discounted Sales: Every brand has a standard market operating price( which could be MRP in few categories). And brands also have some sales through consumer discounts which are over and above the MOP Discounted Sales Contribution= Sales Volume with Discounts/ Total Sales Volume If you have a strong brand, the contribution of discounted sales will be lower. The ability to have more sales at the market operating price is a sign of a strong brand 3. Performance Ads Driven Sales: Every brand will have some organic sales( brand searches, repeats, Marketplace SEO etc) and some paid sales( Amazon ads, Google/FB ads) Performance Ads Driven Sales Percentage= Sales due to ads/Total Sales If you have a strong brand, the contribution of ads driven sales will be lower. A strong brand has higher repeats, higher brand searches and rank organically on top for generic searches on marketplaces 4. Performance Ads Driven Visitors: On the D2C website as well as marketplace listings, brands get both organic( brand searches and SEO) and paid ( Amazon Ads, Google/FB ads) visitors While the previous metric of ads driven sales is difficult for overall attribution( people clicking on ads to come to D2C website buys organically from marketplace is common), this is a easier metric to calculate Percentage of Performance Ads Driven Visitors= (Ads driven visitors on Marketplaces+ Ads driven visitors on D2C)/ (Total Visitors on Marketplaces+ Total Visitors on D2C) As brand strength grows, percentage of ads driven visitors should keep reducing 5. Share of Spends/Market Share: Share of spends in a category is the marketing spends done by the brand as a percentage of spends done by the entire category in a year. If a brand has a higher market share than share of spends, it means 2 things - Higher Conversion Rates & More Efficient Marketing Engine - High Baseline Sales When brands start, they would most likely have higher SOS than market share( as baseline is 0). But as brand strength grows, this number should be lower Strong brands should result in strong businesses. Done right, Investment in Brand Building always pay off financially. It means stronger brands are less reliant on performance marketing, discounts and can increase prices without drop in volumes.

After 9 years in marketing, I've finally come to a conclusion. This is the 1st marketing problem of B2B businesses: Skipping crucial steps to chase quick results. Here's what it looks like in practice: - Running outbound campaigns without a target market - Building a CRM routine without revenue objectives - Publishing content without a desired positioning - Writing a homepage without a clear messaging You understand the pattern: Doing (marketing operations) without (strategic input). It's like choosing the walls' color while your house doesn't have foundations yet. And all the marketing problems stem from this lack of clarity. - How can you focus without knowing what to achieve? - Why would prospects listen if you don't know what to say? - What will you say if you don't know who you're speaking to? Now here's the truth: Clarity doesn't come from a magic illumination. You get clarity by answering the right questions, one by one. Here are the 4 levels of B2B Marketing to help you: 1. Always start from Business Strategy Marketing must be aligned on the objectives and revenue targets of the business. Otherwise, you're building an art & craft department. 2. Turn it into a Marketing Strategy When the business objectives are clear, a great CMO will define a 'game plan' on how to achieve them. This step is the combination of strategic decisions (positioning, messaging) and planning (roadmap, budget). 3. Choose Marketing Tactics accordingly With the strategy in place, the next step is to choose actionable marketing activities. They need to make sense based on the target audience and desired positioning. 4. Run Operations to apply those tactics Here, it's not about answering questions anymore. It's about consistently executing marketing tasks and iterating based on results. You need the right skills, discipline, and clarity on what to do. *** Follow me Pierre Herubel for daily marketing tips. Join my free 5-day email marketing course (click "visit my website")

Paid media is often the largest expense on your P&L. It’s easy to lose sight of this and chalk up the "attribution wars" (as they are known on Twitter) to a silly debate or a waste of time, but a few % improvement on your paid media efficiency might be worth millions (or tens of millions depending on the size of your business). It’s a worthwhile investment to try and cut through all the noise and to find “true north”. But we see so much confusion around the best tools and approaches. For me, it boils down to a few core principles: 1. Be skeptical of anyone who has a financial incentive to deliver good news or grow your ad spend. When I first joined Netflix and ran controlled experiments, it felt like the red pill moment in The Matrix. So much of the reporting you see out there from vendors and (some) agencies is more of an illusion than reality, with all of them taking credit when your business is doing well. 2. Search for the true *causal* effects. Incrementality is such a big buzzword now and it is often so misused that it doesn’t mean anything. Attribution is built on correlation, and has become even less reliable since ios14. Experiments establish causation by introducing a control group so you can see what would have happened anyway (think of RCTs in healthcare). Since you still need a more real time view of performance, use experiments to calibrate your day to day attribution. 3. Prioritize scientific rigor. I can’t tell you how important it is to sweat the details - there is so much you can miss if you don’t have a background in data science and statistics. Re: vendors who offer all in one solutions or claim they have solved this problem with a “magic” pixel or something, dig in. Achieving both accuracy and precision in marketing measurement is very, very hard. We work with PHD economists and world renowned professors on this stuff and even they will say that marketing measurement is extremely difficult for a whole host of reasons (economists reading this, please chime in here!). I’ve linked some more objective 3rd party resources in the comments for those just starting on the journey.

Don't let Meta pick your winning ads Instead do this: What Meta does well - Meta optimizes toward ads that drive conversions. - If they make you more money, you spend more money. What Meta doesn't do well: - Their reporting has cookie loss. - They struggle to attribute top funnel ads that convert days or weeks later - Optimize for business outcomes, e.g. Retail lift. Amazon spillover. LTV. Which means Meta can make bad decisions: - Kill an ad driving retail purchases they can't measure. - Scale an ad with great engagement but terrible downstream economics. And I've seen the bigger and complex your brand is beyond Meta and a single conversion. E.g. retail, Amazon, subscription, varying gross margins across products The bigger the gap between what Meta optimizes for, and what outcome you actually want. This is exactly how we pick our winning ads: 1. Set up one ABO for testing and top of funnel scaling 2. Set up one ASC per product for additional scale 3. Use a multi-touch attribution tool focused on new customer CPA (not Meta's reported CPA) 4. Watch the difference between total clicks, first touch and last touch attribution to know which ads drive top of funnel vs bottom of funnel 5. Check the predicted LTV and AOV by ad. An ad with a higher CPA but 2x the AOV or stronger rebill rate might be your most profitable ad. 6. Plot all ads with spend on X axis and CPA on Y axis (adjusted for these assumptions) 7. Draw a Poisson confidence curve at 90% confidence against your target CPA 8. Any ad past that curve with enough spend? Cut it. 9. Any ad below target CPA but barely spending? Underscaled. See if forcing spend drives performance. 10. Budget changes: 20% at a time. Big swings push ads into relearning. 11. Check twice a day at $150k+/month spend. More smaller moves always wins. 12. Scale winners into new placements first, then increase budget 13. If one ad monopolizes spend in a campaign, duplicate the campaign, pause that ad in the duplicate, lower budget on the original Meta's goals are not your goals. Meta does not have as much data about your business as you do. How are you handling this in your accounts?

I think we’re measuring the wrong stuff… and it’s quietly killing momentum. 2026 has to be the year we fix it. Impressions. Clicks. MQLs. “Engagement.” The real game is happening in DMs, Slack threads, forwarded newsletters, and meetings. Here are 6 metrics I’d focus on in 2026 GTM (and why they matter). 1) Conversations → conversions What it is: Of the conversations your content starts, how many turn into a real next step (intro, meeting, opp). Why it matters: Content doesn’t “generate leads.” It generates conversations. Pipeline comes from what you do next. How to track: Tag every inbound convo (DM/email/reply) and mark the outcome: no fit / nurture / meeting / opp. 2) REAL ICPs engaging with content What it is: Not “engagement.” Engagement from the right people (titles, seniority, company tier, intent). Why it matters: 1 CFO at a target account > 1,000 random likes. How to track: Maintain an ICP list (titles + account tiers) and measure: % of engagers who match ICP of target accounts engaged per week repeat ICP engagers (X touches in 30 days) 3) Brand mentions inside ICP-relevant conversations What it is: How often your brand comes up when your ICP is discussing the problem you solve (not when you post). Why it matters: This is the difference between “content that performs” and a brand that gets recommended. How to track: Collect signals: customer calls (“we heard about you from…”), community moderators, partner chatter, dark social screenshots, and sales intel. Even a simple monthly “mention log” works. 4) Conversation velocity What it is: The speed from publish → first qualified conversation, and from convo → meeting. Why it matters: Velocity is the earliest indicator your messaging is landing. If it’s slow, you’re not sharp enough yet. How to track: time-to-first-ICP-convo after a post/report time-to-meeting after first touch “conversation depth” score (comment → DM → problem share → meeting ask) 5) Brand + category position What it is: Are you being associated with a clear “lane” (category/point of view) or just “a vendor who posts”? Why it matters: In 2026, positioning is distribution. If people can’t summarize your POV in one sentence, you’re invisible. How to track: Quarterly “message recall” check: ask prospects/customers: “What do we do?” “What do we believe?” “What are we known for?” 6) Dark social + word-of-mouth What it is: The off-platform sharing that actually drives deals: forwards, screenshots, Slack drops, “my friend sent me this.” Why it matters: A huge percentage of B2B buying happens in private. If your GTM can’t see dark social, you’re flying blind. How to track: “How did you find us?” (mandatory field) inbound screenshots / Slack mentions private replies after posts If your 2026 GTM dashboard doesn’t include conversations, ICP quality, dark social, and category position, it’s going to keep optimizing for attention… while someone else captures intent.

A fashion e-commerce company CMO asked me this question. They recently launched a 360-degree brand campaign and promoted only one particular high-frequency category item in the campaign.   They want to understand how they should define the campaign's success metric. My answer was this. First, define what success metric should answer to state that the campaign is successful. - Is the sale of items on the platform increasing? - Is the intent to buy the item increasing on & off the platform? - Is the promoted item the right one to bring new users to the platform? - Is the campaign sustainable? Then, define the metrics to answer the above questions exactly or directionally. 1. The metric to answer the first question was simple. - Week-on-Week & Month-on-Month growth trends of the product will answer this. - This should be measured at two levels but we will see this in the fourth question The metric used to measure the second question is two parts. 2.1. Measuring items intent on the platform  - Number of searches( including spelling mistakes) for the items  - CTR of the item banners/highlights on the homepage¬† 2.2. Measuring the items intent off the platform   - Brand + items search volume (Google, Bing & Marketplaces)   - Brand search volume (Google, Bing & Marketplaces)   - Engagement rate of items specific posts on Social media The third question is important to understand the potential of the item. 3. Is the item being promoted the right? - Number of new users growth trend pre-, during, and post campaign - Retention/repeat purchase rate of promoted item bought/searched user The last question needs to be measured with a slightly longer-term lens. 4. Campaign sustainability   - Overall sales lift during the campaign   - Overall new user lift during the campaign   - Overall sales baseline shift after the campaign   - Overall new user baseline shift after the campaign Can you add one more question and its metric to polish this further?

Here's a surprising trend in marketing: More measurement. More confusion. MMM, attribution and experiments often disagree. Marketers are left guessing what’s true. I recently asked a room of 45 marketers to sum up measurement in one word. Their top answer? Confused. New Kantar research backs this up: - 55% of marketers have seen their own data contradict itself. - Advertisers now use 3.8 measurement tools on average. No wonder things don't always align. Faced with contradiction, many just fall back on the  default method at their company, even if it’s badly flawed. "We all know it's wrong but they're the KPIs we're judged on." This isn’t about picking a favourite. It’s about building a system that works when your tools don’t agree. Here’s some thoughts on how it can be done: 🧭 5 Moves to Make Measurement Work 1. Move from a "Source of Truth" to a "Suite of Truth" No tool gives the full picture. A suite lets you calibrate multiple angles and reduce blind spots. 2. Make Incrementality the North Star The goal isn’t to track everything. It’s to measure what actually drives growth you wouldn’t have got otherwise. Lift not last-click. 3. Think Incrementality Ladder Top - Experiments, test and control, the gold standard  Middle - Modelled data like MMM  Bottom - Attribution for quick but non-incremental signals Don’t default to the bottom. 4. Run Experiments Alongside Modelling Test-and-control studies remain the gold standard. Use them regularly to validate, calibrate and strengthen your models. 5. Get More Critical of Measurement Improve data quality. Challenge assumptions. Audit methodology. Connect directly to the bottom line. The Suite of Truth will look different for every organisation. But the goal’s the same: clarity, not convenience, incrementality, not information. Be great to hear how you're tackling this.  When your MMM, experiments and attribution disagree, how do you turn the swirl of confusion into something you can actually use?

For half a decade, I thought I was tracking the right metrics I was wrong Revenue. Growth rate. ROAS. Conversion rate. New customers. Repeat revenue All important But they could tell me the business was growing without telling me whether that growth was making the company more valuable You can buy more traffic, discount more aggressively, and acquire less-profitable customers while the top line keeps going up The business gets bigger That doesn’t automatically mean its equity value does A stronger Brand should make future revenue easier to earn, more profitable, and less dependent on buying every sale Here are the 11 metrics I wish I’d started tracking sooner, framed as questions: 1. Are branded organic searches growing faster than revenue? 2. Are contribution dollars and contribution margin going up? Contribution Dollars = Revenue - variable costs like COGS, marketing, and shipping 3. Is direct and branded search revenue growing faster than overall revenue? 4. Is the gap between gross and net sales shrinking? This signals less reliance on discounts and fewer returns 5. Are 30, 60, and 90-day incremental LTV going up, excluding the first purchase? 6. Is reach growing as fast as—or faster than—revenue? 7. Have your worst days gotten better? One way to measure this: is the average of your 30 lowest-revenue days trending up? 8. For organic search, is revenue per session rising while sessions are growing or stable? 9. Is your share of branded organic searches growing versus your competitive set—at both the Brand and category level? 10. Is Baseline Revenue growing, both in dollars and as a percentage of total revenue? I define Baseline Revenue as revenue from direct traffic, organic search, and organic social referrals It’s imperfect. But if it’s rising in dollars AND as a percentage of revenue, good things are generally happening 11. Is Baseline Revenue per branded organic search going up? Branded searches are an imperfect proxy for the Brand you’re building. Baseline Revenue per search shows whether you’re monetizing it better If searches are soaring but Baseline Revenue per search isn’t, that’s something to audit — A few caveats: None of these metrics are perfect. You can game any of them They’re also mostly leading indicators—not the ultimate company scorecard The ultimate outcome is more operating profit and net cash over time The right metrics also change with the company’s stage, economics, and strategy. A five-month-old company shouldn’t use the same scorecard as a 100-year-old company But if you can honestly answer “yes” to most of these questions, there’s a good chance the quality of your growth is improving And that gives you a better chance of building a more valuable company—not just a bigger one Question for the people of the internet: What else do you track to understand whether growth is increasing the quality and equity value of the business?

Your campaign grew 30%. Your actual return? 2%. I ask this one question in almost every interview. "How do you measure the return on a marketing campaign you ran?" And almost every time, I get some version of "we tracked impressions and engagement" or "sales went up during the campaign". A very superficial assessment of the campaign. Here's why that answer worries me. If you don't know what actually worked, how do you decide what to scale? A typical campaign runs TV, digital, print, radio, and influencer collabs simultaneously. Can you really isolate what drove the result? Not really. Large brands use Market Mix Modelling (MMM) to understand the impact of multiple inputs, such as trade promo vs consumer promo vs pricing, on sales revenue. If you are a small brand, you can build in layers, start with one medium, measure impact, then add another. The sum of two should be greater than one. If it's not, something isn't working. Studies by Les Binet and Peter Field show that adding an incremental medium should improve performance. Next, pick a control market and a sample market. Run your input in the sample, keep the control clean. No promos, no other activity running there. And this is where people slip up. Your control can't be a market where some new trade scheme or promo is launched; that's not a fair comparison. Ideally, both markets should be at a similar maturity. Comparing a city at 10% penetration with one at 5% is not a clean read. The magic word when looking at success is incrementality. If your sample market grows by 30% and your control market grows by 28%, your true return is 2%. Not 30%. That 28% would have happened anyway. This one reframe changes how you evaluate spending completely. Of course, not every situation allows a clean geo-split or audience-split A/B test. In those cases, you have to resort to a simple pre vs post. Conversion was 10% before, now it's 15%, delta is 5%. And for the more evolved version, you build a predicted baseline using historical trendlines for the sample market, then compare actual growth against what would have happened had you not invested. This is especially useful when you genuinely can't find a decent control market. Knowing what worked and scaling it up can significantly improve returns. What's your go-to method for measuring whether a marketing input actually moved the needle, or do you find most teams still rely on gut? #marketingroi #incrementality #experimentdesign #brandstrategy #marketingmeasurement PS: Views expressed are personal