Surf Excel Ad Testing Case Study: Does Vintage 'Pour Rub Pour' Beat 'Daag Acche Hain'?

A premium leader's dilemma, two creative territories. 'Daag Acche Hain' is one of Indian advertising's most recognised purpose-led narratives: it reframes stains as evidence of a child's growth and positions the parent — not the dirt — as the hero. That emotional territory is central to why the brand sustains premium pricing, because meaning, not cleaning power alone, is what competitors cannot copy overnight. 'Pour Rub Pour', from the vintage 2016 campaign, sits at the other end of the creative spectrum: a tactical, functional hook that turns product usage into a memorable demonstration. One builds long-term brand meaning; the other drives short-term provocation and usage salience. The research question this study was designed to answer is precise: should Surf Excel re-integrate tactical functional hooks into its long-term purpose-led narrative — layering the two — or keep the territories separate?

Why guessing is expensive for a market leader. When a brand owns premium pricing, every creative pivot carries asymmetric risk. Lean too functional, and you risk dragging a purpose-led premium brand into commodity comparisons where the fight is about price and promotion — a fight leaders usually lose margin on. Stay too abstract, and you risk losing the tactical punch that drives trial, usage salience and shelf conversion in a laundry category where liquid-detergent adoption and washing-machine ownership concentrate in the metros. Add media inflation across TV and digital, and a wrong creative bet across six metro markets is not a mistake you quietly absorb — it is a crore-level lesson. Kya scene hai, then? The honest answer for a leader: you do not guess between heritage and strategy; you design a study that measures both, untainted, before a single rupee of media is committed.

Turning a boardroom debate into a testable design. This is where ad testing earns its budget. Instead of arguing about which territory 'feels' right, the study frames the tension as measurable constructs: emotional response, functional persuasion, brand fit, head-to-head appeal and purchase intent — captured per creative, per cohort. Because the creative-diagnostics findings were pending fieldwork at report generation, what this case study can already show you is equally valuable: how to structure the question, the design and the sample so the eventual data is decision-grade. Teams studying adjacent tensions — heritage versus reinvention, purpose versus performance — will find the same architecture applies, whether the brand is a detergent giant or a challenger mapping switching behaviour like the brand switching case study from India. The pattern is universal: define the tension, isolate the creatives, segment the memory, and let evidence arbitrate.

Why side-by-side comparison corrupts the read. Show a consumer two ads back-to-back and ask which is better, and you get a verdict — but a contaminated one. The first ad frames the second; contrast effects, fatigue and primacy quietly rewrite every score. A pure monadic design, where each respondent sees only one ad, gives a clean read but multiplies fieldwork cost with every extra creative and removes within-respondent comparability. The study chose the middle path with a twist: a sequential proto-monadic design with randomized order. Each respondent experiences one ad completely first — story, emotion, functional message — with that initial emotional and functional response captured untainted, before any direct comparison happens. The design then proceeds to the next creative, and because ad order is randomized across the sample, whatever sequence effects remain are distributed evenly across cells rather than piled onto one unlucky creative. You get isolated, untainted per-ad reads and a fair head-to-head — ekdum solid methodology for a two-territory fight.

Unaided first, aided second: the diagnostics battery. The questionnaire sequences judgment deliberately. Spontaneous reactions — including per-ad eye-catching ratings — are captured before any prompt can anchor them, preserving the gut response that predicts real-world attention and recall. Only then do aided diagnostics come in: brand fit (does this creative belong to this brand?) and head-to-head appeal once both creatives have been experienced. The battery also builds in brand-association co-occurrence analysis, which maps which associations travel with which creative territory — the raw material for understanding whether functional hooks drag down, or genuinely complement, purpose equity. Every instrument is chosen so the eventual read-out answers the actual business question, not just produces a scorecard nobody can act on.

The statistical spine: paired t-tests by cohort. Purchase intent — the metric finance actually cares about — is designed for a paired t-test comparison between the younger and older parent cohorts (25-35 versus 36-48). Pairing matters: because the same respondent framework evaluates both creatives under randomized order, the comparison controls for individual response style rather than comparing two different groups of people. Combine that with cohort segmentation and you can see not just whether a creative territory wins, but whether the win differs for parents who lived the 2016 era versus those who did not. For the full platform logic behind this kind of build — stimulus handling, routing, quota enforcement and analysis — see ad testing platform India and the creative-survey layer in ad creative testing survey India. The design is the finding: most contested creative debates die the moment the measurement is honest.

N=200, exactly as targeted. The study set a target of 200 completed responses and achieved 100% of it — no shortfall, no quietly widened error bars. Every respondent is a parent in a Tier-1 metro who participates in laundry and grocery decisions for the household. That last filter matters more than it looks: in a category where the purchase decision is shared, sampling only self-declared primary shoppers would have skewed the read toward one half of the household. And note the wording — the completed sample is the verified sample. Raw fieldwork numbers were deliberately larger, and the next section explains why half of them did not make it into the final dataset.

A 50/50 gender split that mirrors real households. The sample holds 100 men and 100 women — a deliberate 50/50 split reflecting joint grocery decision-making in double-income urban households. In metros like Mumbai, Delhi, Bangalore, Chennai, Kolkata and Hyderabad, laundry decisions are negotiated between partners who both work, both shop and both watch the advertising. A single-gender sample would have answered a question this study never asked. Gender balance here is not a compliance checkbox; it is a modelling decision that lets the eventual analysis read the whole household, not one convenient half of it. Any detergent brand study India runs should start from the same household logic.

83% in the core parenting band, 17% who remember 2016 first-hand. Fully 83% of respondents fall in the core parenting age of 25-45 — 41.5% aged 25-35 (n=83) and 41.5% aged 35-45 (n=83) — two evenly weighted cohorts that make the generational comparison statistically meaningful rather than anecdotal. The remaining 17% are aged 45-55 (n=34, capped at a quota of 48): the parents who experienced the original 2016 'Pour Rub Pour' launch first-hand, as adults managing their own kitchens and washing machines. Their nostalgia — or fatigue — with vintage creative is exactly the variable the study isolates. That is what separates a cohort design from a demographic afterthought: the age bands exist because memory of the 2016 era is a research variable, not decoration.

100% Tier-1 metros, on purpose. Every respondent lives in Mumbai, Delhi, Bangalore, Chennai, Kolkata or Hyderabad — a deliberate full-metro design that matches where liquid-detergent penetration and washing-machine ownership concentrate. Premium liquid detergent is, to a first approximation, a metro product; testing its creative in markets where the product-market fit does not yet exist would contaminate the read with category confusion. When you replicate this architecture for your own brand, start from the same logic: define the geography where the decision actually happens, then recruit there. The panel foundations in verified panel India and the wider context in Indian consumer market research both follow this principle.

The 50% cut, explained without embarrassment. The study fielded to 400 raw respondents and excluded 200 of them — a 50.0% quality exclusion rate — after attention checks and data-quality screening. What does that screening actually catch? Speeders who race through a video-ad questionnaire faster than the stimulus can possibly play. Straight-liners who tick the same rating for twenty consecutive questions. Respondents who fail attention checks designed to separate readers from skimmers, and low-quality completions that would otherwise sit in your dataset like watered-down milk in the analysis. The result: a final sample of 200 completed responses where every single record survived deliberate hostility testing. For a study measuring subtle things — nostalgia, brand fit, purchase intent — that dataset is worth more than 400 unverified completes, full stop.

Why a brutal exclusion rate is a feature, not a bug. Here is the uncomfortable industry secret: cheap clicks are cheap because nobody screens them. A 50% exclusion rate sounds alarming until you understand the design logic — the team over-recruited at a deliberate 2:1 ratio precisely so that aggressive quality screening would still land the target n=200 at 100% achievement. That is what high-integrity panel work looks like: you plan for the share of responses you will refuse to pay for. Panels built on verified identities, like the 20M+ ZK-verified zero-bot SuperJ app community spanning Tier 1/2/3, start from a higher floor — but the screening discipline stays the same, because verification at entry is not a substitute for quality checks inside the survey. Decision-makers should ask any vendor one question: what was your exclusion rate, and what did you exclude for? If the answer is 'we don't exclude anyone', the data is not cheap — it is expensive later.

An honest note on what this report does and does not contain. This case study documents the full study design, sample architecture and quality framework — the parts that were locked before fieldwork. The creative-diagnostics findings — the eye-catching scores, brand-fit ratings, head-to-head appeal and purchase-intent results — were pending fieldwork data at report generation; data collection was the required next step, and the report says so plainly. We would rather publish an honest framework than a fabricated verdict, and we encourage you to be equally suspicious of any 'case study' that never shows its sample or its screening. For the machinery behind this discipline — attention checks, speeder detection, straight-lining flags and ZK dedup — see survey data quality India and the panel foundations in verified panel India.

Choosing your design: proto-monadic vs monadic vs paired comparison. Pure monadic testing — one ad per respondent — gives the cleanest isolated read but needs large samples per cell, which multiplies cost with every additional creative. Paired comparison — forcing a choice between two ads in one sitting — is fast and statistically neat but bakes contrast effects into every score. The sequential proto-monadic design used here sits deliberately in between: it captures untainted per-ad responses first, then allows direct comparison, with randomized order neutralising sequence bias. Choose it when you are comparing two or more creatives on both emotional and functional dimensions across cohorts — the exact situation in this study. Plan your quality headroom too: this study recruited at 2:1 and still hit its target after a 50% exclusion rate, which is a sensible norm for video-stimulus research with hard attention checks.

The run-your-own playbook on Hercules Works. One: write the business question before the questionnaire — 'should we re-integrate functional hooks into our purpose narrative' is a question; 'test these two ads' is not. Two: load your creatives through proper stimulus management so video playback, order randomization and rotation are enforced by the platform, not by hope — see survey stimulus management India. Three: define cohorts that map to real memory and behaviour, as this study did with its 25-35 versus 36-48 parental cohorts. Four: field to a verified panel — the SuperJ app community of 20M+ ZK-verified users across Tier 1/2/3, zero bots — and screen ruthlessly on attention checks. Five: let Poseidon, the AI analytics engine, run the diagnostic battery and the paired significance tests. The end-to-end build logic is documented in hercules survey creation process India, and the platform story in ad testing platform India.

The pricing ledger, so you can budget today. Hercules Works pricing is built for Indian marketing budgets, not New York procurement: the Free plan is ₹0/month — permanent, not a trial — with 10 AI research chats, access to 100 SuperJ users and 3 campaigns. Starter is ₹1,119/month, or ₹895/month billed annually with 20% off. Pro is ₹30,000/quarter, or ₹24,000/quarter on annual billing. Every new account also gets 100 responses free in the first month, so you can pilot a two-ad test before committing a rupee. Compare that with legacy ad-testing programmes priced in lakhs per wave, and the value case writes itself — paisa vasool is an understatement. Trusted by Unilever, Kantar, Govt of Karnataka, ICICI Prudential and SBI Mutual Fund, the platform is proven at enterprise and startup scale alike. Start free, put the vintage-versus-modern question to your own consumers, and let evidence settle what boardrooms cannot.