Marketing Campaign Evaluation

Attribution is overrated. Incrementality is what actually matters Every new-age brand wants to know what’s working. Meta ROAS is looking good. CAC is steady. Revenue is growing But here’s the truth: Your Meta ad might get the conversion. But did it cause the conversion? That’s the difference between attribution and incrementality. Most dashboards, attribution tools, and agency reports stop at attribution. But if you’re a brand selling across Amazon, Flipkart, GT, MT, Q-com, and D2C—pure attribution will always lie to you Because the sale might happen on Amazon. But it might have been nudged by a Meta video or a YouTube bumper ad 4 days ago. You don’t need a full-blown Marketing Mix Model to get started. There are simpler, street-smart ways to directionally understand what’s working—and what’s not. Here are 4 that have worked for us at Atomberg: 1. Geo Split Testing Pick two similar markets. Run campaigns in one. Don’t run in the other. Then track: • Branded search volume • Sell-through on marketplaces • Secondary sales from GT counters If the test market moves faster than the control, you’re seeing true lift. That’s incrementality. 2. First-Time Buyer Growth vs Returning Buyer Growth Track whether your growth is coming from first-time buyers or repeats. If your campaigns are just bringing back old customers—you’re not creating net new demand. But if there’s a spike in new buyers across Amazon, Flipkart, D2C—your campaigns are likely working at an incremental level 3. Paid Traffic vs Organic Trend Lines If paid traffic, clicks and spends are going up—but your organic sales or branded search isn’t moving—you’re likely just harvesting demand that already existed. But if organic lifts alongside paid—your ads are creating interest. Not just closing it. Directionally, this is one of the simplest sanity checks most teams ignore. 4. Channel Crossover + Offline Signal Mapping Your Meta ad may not show up in last-click attribution. But it might have nudged the consumer to visit your store or buy on Amazon. You can detect this through: • Post-purchase surveys (Where did you first hear about us?) • Branded search + store footfall spikes in campaign-active cities • And most powerfully—offline signals passed back to Meta At Atomberg, we pass back data from installations and warranty registrations—including pincode and purchase timelines Sometimes, we’re even able to identify this at a unique customer level through their cookies for warranty registration This has helped us understand true incrementality of perf marketing campaigns even for offline sales If you’re only measuring ROAS, you might scale what’s only taking credit for sale about to happen anyway If you chase incrementality, you’ll scale what’s working. For more details, read the full post- link in first comment.

2022's Come and Say G'day campaign from Tourism Australia was a big success. Testing from System1 rated the work in the top 1% of all campaigns tested. Consumers that recalled the ads had a 15pt increase in awareness and consideration. And in the target countries were the campaign ran consideration gains and share of search increased significantly versus the other markets. International tourism spend in Australia now significantly exceeds pre-Covid levels and will land about $7 billion ahead of target this year. No wonder the work won five Effies. But that's a problem for Susan Coghill the CMO. She is a scholar of advertising effectiveness. You can see it in her work. Emotion. Codification. And maintaining the same campaign for years to ensure it has time to work. Most marketers pull campaigns and replace them long before they can have their maximal impact. Susan knows better. Again the data confirms she is right. After two years the same creative tested just as well (in the UK) and actually improved over time in the US. So why change it? Winning ads stay winning ads for a very long time. They usually improve. Why not save creative costs and keep running the same work? Why take the risk? It's here that Coghill makes some interesting points. First, she does not disagree with 'baking your cakes for longer'. Three years of the same creative is an eternity in the tourism and travel category and contrasts with Tourism Australia's former approach of a new campaign each and every year. It's also not a new campaign. Again, unlike less well trained marketers, Coghill is not throwing the consistency bathwater out with the new baby creative. She bills this work as 'Chapter 2'. Same codes, same theme, same fluent device. Fresh but familiar. New campaigns, especially extensions, eventually make sense. The whole media caravan that surrounds new work and launching can positively impact the market. This new work also keeps internal customers, retail partners and employees happy and invested. And finally, as effectiveness scholars there is the tempting ability to apply all the learnings from the first campaign to its second chapter. In the case of Come and Say G'Day II - the realisation that the creative will work better if local stars from China, Japan, America, UK, India and US are featured in each country's messaging - see the amalgam of the new work above. I guess the point is that most brands should maintain their campaigns for much longer than they do. Two or more years makes effectiveness sense and contrasts with the creative myopia of 95% of the industry. But after 3 years there is a case to be made for new work, providing it follows in the footsteps of what preceded it and builds from the lessons and limitations of the earlier work. So new cakes do eventually make sense. But only after the old ones have been baked for long enough. And provided today's gateau looks a lot like 2022's pavlova.  #advertising  #branding  #3yearoldpavlova

A rare treat for all marketers on LinkedIn The case study of how not to use data. Yesterday, a group of consumers who could recall seeing the Sydney Sweeney campaign were asked whether they plan on buying American Eagle in the future. Loads of them all said yes. And it was published as proof that the outrage online was overstated because this ad was clearly working. You can see this in The Drum here: https://lnkd.in/eEVJ2Xh9. Ignoring the ethical debate for a second, as whether something is effective doesn't solve whether we should be doing it in the first place. But it's a rare public case study of how bad data can be used to bamboozle marketers into the wrong decisions. An opportunity to demonstrate two simple research principles that can arm marketers to make better decisions. 1. The Rosser Reeves Fallacy. If you ever segment a group of consumers by those who can recall seeing an ad from a brand, the recall group will simply have a lot more brand buyers in it. This will inflate results. If you buy a brand you notice, interact with and remember more of their advertising. This sort of segmentation can be used to create data that might convince but it doesn't reflect the market. 2. Purchase Intent as a useless metric. Thinking to humans is like swimming to cats, we don't want to do it, but we will if we have to. What we try and post-rationalise to explain our decisions, or predict future ones, has nothing to do with what we will actually do. Decisions are made lazily. Influenced by hundreds of associative memories. Nudging us. This is why creative testing has got a lot better, according to IPA (Institute of Practitioners in Advertising) data, emotion is a far better way of predicting what someone will do compared to persuasion metrics like purchase intent. This feels like marketing 101. It's important. It's how we make decisions as marketers. So my column this week dives into how to spot sloppy insights and how to find better ones. https://lnkd.in/ehy2PXTj PS. I love The Drum and the team there, and anyone looking for a data-backed POV instead of going off marketers' opinions and viral outrage. This is just a good opportunity to assess some public data. Good on John McCarthy and Gordon Young for letting me write this week, arguing a different POV from what they've published. I share #advertising and #marketing insights daily, follow for more.

Nothing sells better than experiential marketing! With so many options available, consumers want something new and there is nothing more engaging than being able to experience and engage with a brand even before you purchase it. Unlike traditional advertising, which focuses on awareness, this strategy creates immersive experiences that keep the audience engaged for long. It is similar to the difference between watching a movie trailer and actually being at the theatre. This is why I feel it will never go outdated -  → Emotional Engagement - Memorable experiences create strong emotional connections, making consumers feel part of the brand story. Something very similar to Coca-Cola’s "Share a Coke" Campaign. The brand used personalized Coke bottles to engage the audience and help start a new bond. People ended up buying not just a soft drink but finding new friendships altogether. → Loyalty - By giving consumers a chance to engage with the brand, you build a cycle of repeat purchases and loyalty. Fevicol tried it in a very fun way. During festive seasons, Fevicol set up repair booths in local markets, offering free fixes for broken furniture. It was a great way to create an experience that resonated with their brand message.  → Global Reach - Virtual activations give you access to new audiences no matter where they come from. Zomato does it very well. It is no longer just a food delivery app but has created a multi-city carnival that brings together music, food and entertainment. It was a feast for different age groups and gave Zomato a presence in people’s lives.  Experiential marketing is honestly one of the best strategies for brands that dare to go beyond the screen. And the earlier you realise that experience matters more than ads, that is when things will start changing for your brand. I have been working with brands across industries and I believe the brands that win will be the ones that don’t just tell their stories but allow consumers to live them too.  Which brand do you think does the best at this strategy for its audience? #influencermarketing

“But what’s the ROI of brand?” A classic and a very very very fair one. Brand, Branding and marketing are full of BS. And for many who came out of business school, brand feels like fluff. I get it....it can be hard to fit all of it into a CFO-friendly Excel sheet. But before we dive into it: Brand Branding are not the same. Brand is why you matter. It’s trust, credibility, perception and ultimately preference. Branding is how you build that perception: tone of voice, design, website, behaviour, experience, etc... Meaning: → Brand is what people think and feel. → Branding is how you make people think and feel the right things. Let’s talk ROI, External first. → +0.7% market share growth per 10% above-share-of-market brand investment. Meaning: Outspend competitors in mental availability, and you’ll eventually outgrow them. Be memorable not just visible. → 80% of B2B buyers will pay more for a trusted brand. Meaning: You stop selling on price. Trust lets you charge more. Price on perceived value (perception), not product or features. → Up to 50% lower CAC. Branded search terms convert better and cost less. Meaning: Brand makes your funnel more efficient. Less convincing. More $. → 2400% ROI in LTV. Meaning: In one case, every $1 spent on brand returned $24 in customer value. (Modelled scenario, not a promise.) → 5% increase in retention = 25–95% profit lift. Meaning: Brand trust = loyalty. Loyalty = recurring revenue and stability. → 64% of B2B buyers say all suppliers look the same. Meaning: If your brand doesn’t stand out, you’re just another line item. Differentiation lives in perception, not product. → Branding is a multiplier, 23% higher revenue for brand-consistent companies. Meaning: Consistency builds trust. Trust drives conversion. This is sales. → 78% of buyers choose from brands they already know. Meaning: If they haven’t seen you before, you’re out, no matter how good your product is. →Brand as Insurance: Strong brands recover faster from downturns. Meaning: When the market shakes, trust keeps you chosen. Brand = buffer. Aka Margin of safety. Let's look at internal ROI (Often missed.) → 43% lower cost-per-hire. Meaning: A strong brand brings in talent not just clients. No pitch needed. → 28% lower turnover. Meaning: People stay. Culture stabilises. Institutional knowledge compounds. → 17% more productivity, 21% more profitability. Meaning: Brand-aligned teams perform better. Purpose powers effort. → Brand clarity = operational efficiency. Meaning: When teams are aligned, they move faster. Less friction. More flow. Costs down, speed up. Sooo: You don’t build brand to feel cool. You build it to win faster, grow better, and reduce resistance across every part of the business. Brand is the strategic asset. Branding is the system that makes it perform. If you ignore them, you’ll pay elsewhere: - Higher CAC - Lower retention - Price pressure - Slower hires - Lost deals - Weaker teams (All data sources in the comments.)

I get this email once per quarter: From: CEO Subject: FW: Sponsorship Opportunity "Should we do this do you think?" [followed by a pitch to sponsor a newsletter / conference / podcast / etc] I remember the first time I got an email like this, and I thought to myself: There must be a better way to answer this question than with my gut feel Here's how I answer it now: Step 1 - calculate a quantitative benchmark I use CPM (cost per 1,000 impressions) If a podcast sponsorship costs $1,000/episode and reaches 500 people (0.5 x 1000), then the CPM is $2,000 I benchmark that against LinkedIn ads (I use our actual LinkedIn CPM from the last 90 days, but if you don't have that number then $50-$100 is a good bet for B2B) Step 2 - do a qualitative assessment (aka gut feel) Now I ask myself - is the $2000 CPM opportunity worth 20x what a LinkedIn ad impression is worth? Maybe... it depends on: -how much more engaged is this audience vs. a LinkedIn ads audience? -does my brand get a credibility boost from being associated with this "publisher" ? -is this audience better than a generic LinkedIn audience because they are more likely to be philosophically aligned with what our brand stands for? -does this audience fit my ICP better than an audience I can build with LinkedIn ads? etc... I frame the question back to the CEO the same way - to get his gut feel too Then, we decide together if it's worth taking a small bet on (decision on whether to scale usually comes later) #b2bmarketing

1 year and 130k orders into TikTok Shop, here's what I've been surprised to learn. 1. Products must be new, in season or niche: Videos must be interesting or relevant to go viral. Our kids products take off during back to school and NFL in Q4. Out of season they are very slow. New eye catching designs perform better than the best seller we've had for years. We had a Winnie the Pooh design go viral striking a cord with niche Pooh community. 2. TikTok Shop sales haven't meaningfully grown TikTok followers: Simple Modern has only added 25k TikTok follower this past year, selling 130k orders with TikTok Shop. Our TikTok follower count grew less this year than it did the year before we sold on TikTok Shop.   Surprising to me considering we've driven 250m+ product impressions. 3. Major halo effect from TikTok to Amazon and Website:   When a product has a successful video driving TikTok Shop revenue, the bump on other eComm channels is clear. We've seen instances of more sales driven by TikTok videos on Amazon + DTC than TikTok Shop.   Customer trust is higher on Amazon and brand's websites.   The real magic is when TikTok videos goose Amazon listing placement permanently. This is a great channel for omni-channel brands. 4. Revenue/video is flat once affiliates have more than 50k followers:   Followers: Revenue/video 0-1K: $13 1k-5k: $25 5k-10k: $40 10k-50k: $75 50+: $100   Affiliates with 50k followers have performed the same as 1m follower accounts. We have not engaged multi-million follower accounts with highly engaged audiences (celebrities). 5. 25% of our sales are "Product Card": Product cards sales are from customers finding product pages from search. They don't pay out commission. 6. GMV Max Ads have multiple benefits: The ads will find videos posted by affiliates and boost their reach. It has been an efficient way for us to spend. We also saw a significant increase in sample requests from high quality affiliates when we started spending on GMV Max. 7. Affiliates asking for 4+ samples are taking advantage of you.:   We've sent 51 affiliates 4+ samples. Only one generated a sale.   13% of our total samples have been sent to grifters. 🙃 8. TikTok Shop is pushing live video: They've told us the algorithm is favoring live content and they expect it to continue. In China Live videos drive most of the sales, which has pushed the US team to lean in. 9. Brand content is a cheat code: Brands that produce content that generates sales are doing the best. Unfortunately that currently isn't us. Not having to pay affiliate commission is very helpful. Unlike affiliate videos, your own content will grow followers and create a flywheel.   ************* TikTok Shop is a uniquely valuable channel since it's also a marketing engine.   It has required a different strategy from us and has been fun to learn. No doubt we are in for a lot of change with the sale to the U.S.   I'd love to read what others have learned in the comments.

AI collapsed the buyer’s journey and made the old playbook obsolete. Teams need to move faster than ever to meet the moment. That’s why HubSpot moved away from the traditional funnel and built a marketing Loop. Here’s the old way: Marketers plan campaigns, execute, and wait to measure performance. By the time teams unlock learnings, the market has already moved. With Loop Marketing, teams run experiments constantly. They build systems that surface signals in real-time to adapt while campaigns run. If a message isn’t perfect, teams course correct. Learnings happen so fast that setbacks never become full-blown failures. Teams that bake evolution into their campaigns outpace the linear approach. Learnings compound, and larger strategies can be optimized continuously. That’s the competitive moat most CMOs aren't building yet. The brands that win have systems that let them learn and adapt. #LoopMarketing #MarketingLeadership #ContinuousOptimization #CMO

We ran a 90-day test turning one client's blog content into Reddit posts, Quora answers, and Medium articles. The results: 285% increase in brand searches. 340% spike in direct traffic. 156% more organic traffic. Here's exactly how we did it: Most companies write a blog post and call it done. That's leaving 90% of the value on the table. We took existing content and systematically repurposed it across platforms where their audience actually hangs out. 1. Reddit Strategy Reddit drives massive referral traffic when done right. But most brands spam it and get banned. Our approach: • Found 8 subreddits where the target audience actively asks questions • Pulled real questions from those communities • Turned existing blog content into helpful Reddit answers • Added value first, mentioned the brand second No self-promotion. Just genuinely helpful responses that happened to reference our client's content. 2. Quora Strategy Quora gets 300+ million monthly visitors. Questions rank on Google for years. We identified high-traffic questions in the client's niche and crafted detailed answers using repurposed blog content. Each answer included direct response to the question, supporting data from the original content, and link back to the full resource. These answers became evergreen traffic sources. 3. Medium Strategy Medium articles get indexed fast and rank well. We took core sections from the client's best blog posts and reformatted them as standalone Medium articles. Key tactic: We targeted slightly different angles of the same topic to avoid cannibalization. Each piece linked back to the original resource for "more details." The Results (90 Days): 📈 Brand searches: +285% (from 50/month to 192/month) 📈 Direct traffic: +340% (from 120 visits to 528 visits) 📈 Organic traffic: +156% 📈 Referral traffic: +420% And this was all from content that already existed. Zero new content creation. Just smart distribution. Why This Works: When people see your brand mentioned across Reddit, Quora, and Medium answering real questions, you stop being "just another company" and become a recognized authority. Those brand searches? People Googling your company name after seeing you help someone on Reddit. That direct traffic? People typing your URL directly because they trust you. Action Steps: 1. Find 5 to 10 of your best performing blog posts (check Google Analytics) 2. Identify where your audience hangs out (Reddit? Quora? Medium?) 3. Pull real questions from those platforms 4. Repurpose your content to answer those questions 5. Post consistently (we did 2 to 3 posts per platform per week) Don't just copy paste your blog post. Reframe it to answer the specific question. Add context for that platform's audience. Mention your brand naturally. Let the value speak first. Track brand search volume, referral traffic from each platform, direct traffic trends, and time on site.

Just got off another audit call. The third one this month where I had to tell a marketer that the thousands they'd spent on LinkedIn Ads over the past month were, basically, 100% wasted. Honestly, it probably would've been more entertaining to light the cash on fire. The culprit? LinkedIn Audience Network (LAN) was left on with no blocklist or allowlist. This case was especially painful. They had healthy daily budgets and were using Max Delivery bidding. When I checked the placement breakdown, only $0.03 had been spent on LinkedIn itself. But many thousands went to LAN. 🤦♂️ (I would've expected Max Delivery to bid high enough to at least win some LinkedIn inventory. Instead, it happily spent the entire budget on LAN traffic at around $5 CPCs, most of which looked like bots or spam. That one hurt.) I love LinkedIn Ads. I've spent the last 15 years building my career around the platform. So when I see things like this, I usually try to defend LinkedIn. I can't defend this one. The platform shouldn't quietly spend your budget on low-quality placements because you didn't know enough to disable a default setting. If you're wondering whether this might be happening in your account, here are a few warning signs I always look for: • News Feed CTRs above 2% (unless you're intentionally running Engagement or Brand Awareness campaigns) • Landing page CPCs below about $6 • Hundreds of website visitors with little or no conversion activity Those three together are often a dead giveaway. If you want to test LAN, I recommend treating it like its own experiment: 1) Run it in a separate adsdet that's LAN-only. 2) Bid it down to around $3 CPC or less. 3) Use a strict blocklist or allowlist to eliminate low-quality apps and publishers. (I shared mine in the resources section of Episode 150: https://lnkd.in/gSMGMrpA) Or, just turn it off. What's been your experience with LinkedIn Audience Network? Have you seen it drive quality traffic, or has it mostly burned budget? #LinkedInAds #B2BMarketing #LinkedInMarketing