Blunt's buyers are city commuters, frequent travellers and gift buyers, and each has a different reason to own a Blunt. We replaced polished studio shots with street-style photography, shot scenarios for each audience, and built tailored ad journeys around that creative. Then we automated delivery around the weather itself: campaigns that activate during rain and wind, when purchase intent peaks, backed by predictive modelling of Blunt's most profitable cities to enter next. Every dollar spends at the moment someone is most likely to buy.

Google and Metas own attribution and reporting isn't reliable. Using years of performance data, we built a Marketing Mix Model for Blunt, backed by forecasting, that measured what each channel actually contributed to revenue and predicted the ROAS for each next dollar of spend. That meant we could growth sales and marketing budgets with confidence and make critical supply chain decisions as we scaled up paid ads across Australia, USA and Europoe.

Our modeling showed Google Ads had more headroom, so we continued to raise spend for the products and markets with the strongest predicted return.

The result: In months, revenue from Google Ads grew 40%, twice the rate of the extra spend, while cost per conversion fell 30%.

Our MMM forecast said Blunt could push spend up 50% before returns flattened, so we did, across prospecting and remarketing. Revenue from social ads rose 43%, and cost per conversion dropped 39%. The result: Our team spending significantly more while conversions get cheaper.

Before the model came the machine it measures. We mapped how customers moved through the Blunt website and where they dropped off, then A/B tested page structures, content priorities and purchase flows against what each visitor needed to commit. Tracking showed exactly which pages, messages and elements moved people to purchase, so the site was redesigned around what was working. Conversion rate increased 301% and website revenue grew 104% in seven months.