Trusting The Twitter
You’ve just booked a flight, bought a coffee machine or reserved a day spa. At some point during the booking process, you’ve accepted the cookies and now it starts: You receive pop-ups, banners, notifications, or alerts for the exact same flight, coffee machine, or day spa that you just purchased.
This kind of product suggestion is annoying and is neither helpful for buyers nor for sellers. In this case, the algorithm was just wrong, and the personalized advertising backfired.¹
Online retailers, above all Amazon, have introduced the principle of product suggestions. After purchase, the customer receives suggested products that buyers with similar preferences have purchased. And the system works. 74% of shoppers click on the suggestion and even 34% buy the product directly. Certainly, online retailers want to take advantage of such sales-boosting tools.²
But what about the customers? Are they fans of product suggestions or are they just naïve marketing victims? The answer is, it depends. More precisely, it depends on whether we talk about product suggestions or product recommendations.
Product suggestions mainly trigger indifference
Whether customers are excited or annoyed by product suggestions often relies upon their mood. If they are in good spirits and in the mood to buy, they rather feel positive about product suggestions. If they just want to place an order and complete the process quickly, pop-ups or banners with product suggestions are more likely to cause frustration. In sum, shoppers usually are indifferent to product suggestions.
Product recommendations instead save time…
Let’s say you want to buy a new washing machine. What are your first steps? We assume you sit down in front of your laptop, open a search engine, and research online which washing machine is the best for your needs. Are we guessing right? If so, you’re part of the majority of the population that goes online to find out what’s on offer and compare products before making large purchases. What consumers consider a large purchase is, of course, an individual state of mind and depends on factors such as income.
But the mass of information to process when looking for the perfect washing machine can also be overwhelming and time-consuming.
So, when we get a credible product recommendation that shortens the selection process, we are grateful and will gladly accept it. ³
…and provide purchase security
A few years ago, buying certain products online, such as mattresses or bicycles was unthinkable. This is because they belong to the category of products that people like to try out, examine in person, or feel and touch before buying. Today, the online markets for mattresses and bicycles are growing considerably. And for exactly this type of product, recommendations are a decisive factor for the purchase decision. Even though buyers can’t lie on the mattress themselves or test ride the bike, they can rely on the evaluation of others who have already used or tested the product. This provides purchasing security and minimizes the risk of regretting the order afterward. ⁴ ⁵
But which recommendations do we as consumers trust most?
With Word-of-Mouth recommendations…
Here we talk about recommendations from family, friends, or acquaintances about purchases that they would instantly make again or that they would dispose of immediately. Such tips from the personal environment are just worth their weight in gold. People who are close to us often have similar preferences. That’s why we intuitively assume that positive recommendations will also appeal to us and vice versa. ⁶
….marketers have a tough time
It is difficult for companies to influence this type of recommendation. However, through convincing product quality and promised features, companies can help to be perceived positively by their direct buyers and thus contribute indirectly to word-of-mouth recommendations.
The success of Influencer Marketing…
For some years now, social media has offered a completely new channel for product recommendations in the form of dedicated influencer marketing. Influencers with a sufficiently large followership present and recommend certain products, either of their own free will or sponsored by companies. The success of such partnerships is easily explained: Influencer fans are the target group reached and lead to sales. Non-followers are often just not reached. ⁷
…mainly depends on trustworthiness
The critical factor is credibility. If followers don’t believe that an influencer actually uses the product and gives it a positive rating, they will not be convinced to buy it. Companies must therefore carefully select influencers to fit their product, e.g. fitness influencers are more promising ambassadors for protein bars than home décor opinion leaders.
A Word-of-Machine recommendation…
Unlike previous tactics, Word-of-Machine product recommendations do not stem from real people, but from a technology-based system based on algorithms. For example, a Netflix series suggested by the system based on previous usage and consumer preferences falls under a Word-of-Machine recommendation. Studies have shown that consumers trust such technology-based recommendations even more than human ones in some cases, e.g., if consumers are interested in a complex product that requires a great deal of advice. Consumers then perceive a scientific recommendation equal to competent advice in a specialty store and prefer it to a subjective recommendation by an uninformed person.
…is on the threshold of gaining full acceptance
However, researchers still see differences in products here. When it comes to functional purchases, consumers are more likely to trust a Word-of-Machine recommendation. In the case of hedonistic purchases, which are pleasure- and emotion-oriented, buyers are more likely to prefer Word-of-Mouth recommendations. ⁸ ⁹
banbutsu is committed to making Word-of-Machine master this threshold
Our mission is to improve word-of-machine recommendations and combine them with the benefits of personal recommendation. We are working on a decision engine that will truly help consumers and simplify their shopping experience. The algorithm is based on the preferences voluntarily shared by consumers and takes into account external factors such as availability, size and weight of a product. banbutsu then only makes recommendations if products are really desired, relevant, and have the suitable characteristics to even be used by the customer. Furthermore, banbutsu considers that recommendations should fit stylistically to the customer and fulfill their visual as well as functional needs.
The development of such a decision engine is complex, but we find it absolutely necessary in e-commerce to give meaningful recommendations and prevent irrelevant suggestions.