Amazon is the largest single eCommerce platform in India and has a spread of audiences from all parts of the country, including an increasing share from Tier 2 and Tier 3 cities/towns. During the winter season, the central, the northern, and certain parts of the southern regions of India see a rise in the demand for water heaters. Racold is India’s most trusted brand of water heaters from the house of Ariston Thermo Group. Unlike the other players in the category (who are home appliances brands) Racold's products are specific to water heaters. As such it often does not enjoy immediate recall as water heaters are a low involvement category. Hence, discovery of the brand especially on eCommerce becomes important yet difficult.

We launched first ever eCommerce campaign that turned misspelled searches into sales. The idea came from the insight that Indians (specially in Tier 2 & 3 cities/towns) tend to type as they speak. As such the keywords they input are often misspelled due to pronunciation used. For example: the word ‘geyser’ is often misspelled as ‘gijar’ and ‘geezar’. Hence, we targeted these misspelled keyword searches made by a newer set of audiences on Amazon by serving search ads in accordance with the vernacular influenced dialect of consumers. This was done in an attempt to ensure that, regardless of the misspelled searches, the consumers found a Racold product that met their requirement and alongside build the brand discovery to a newer set of audiences too.

India is home to 22 major languages written in 13 different scripts with over 720 dialects. This results in Indians being phonetically influenced by their native dialects when pronouncing and spelling common English words, leading to English being mispronounced, misspelled and hence mis-typed. The target audience for Racold was non-metro markets, where the consumers had a strong native phonetic influence on their English usage. We observed that this audience used Amazon to search for geysers (water heaters) wherein their keyword input was misspelled basis their pronunciation. We therefore deployed a phonetically influenced keyword strategy based on the search input of our audience. The strategy was expanded to include keywords and variations basis not just audience input but also regular keywords from the brand, competition and category.

Our insight was validated through our eCommerce insights tool which showed 1.5 MN monthly searches of over 526 misspelt keyword variations. Hence, we targeted the vernacular dialect influenced keyword searches made by a newer set of audiences and showed them ads on a real time basis on Amazon. The pronunciation-based variations were further divided into various ad groups in accordance to preferred products, and keywords were optimized to yield maximum results. Through our tool, we also found out that the miss spelt searches where happening from certain geographies in the country and thus we optimised budgets accordingly. Over the course of 3 months, more than 2500+ keywords variations were used to reach out to newer and current set of audiences using category, brand and competition keywords. Additionally, the misspelt keywords are now a part of all our product listings on Amazon ensuring organic rankings for the brand in the long-term.

In just one month:

• We hit over 1.5 million high-intent searches with 526 misspelled keywords and delivered a 200% increase over planned sales

• We achieved sales worth 33,264 USD with an optimized spend of 2026 USD. This resulted in an overall cost of sales that was brought down from 22% to 6%

• We were able to achieve a 16X return on investment against an industry benchmark of 6X

Additionally, in the first 3 months by using we managed to get sales worth 104,127 USD with a spend of 11,493 USD