Marketing can be a very powerful tool, especially for e-commerce. There are many leads, and it is important to personalise the communication with everyone (which, of course, cannot be done manually). However, if you use this tool carelessly, you can get the exact opposite effect, and users will be shocked. In this article, we prepared for you the list of the most common mistakes in e-commerce marketing that you should avoid. We want to make sure that you know about this in advance so as not to lose all your clients and users.
1. Lack of testing
Sometimes the use of generally accepted algorithms may not help, but harm the store. Even if a practice is considered standard on the market, it is necessary to continue to evaluate its effectiveness on a case-by-case basis. Here’s an example: on the main page of some fashion marketplace, two types of recommendation blocks were tested. In the first case, the user saw a standard section with popular products, in the second – a selection of popular products, adjusted according to his interests. At first glance, it might seem that in both cases, the same correct approaches were used. However, the results of A / B tests showed that where there were “hits of sales”, the size of the average check fell by 15% compared with the control group, which did not see any recommendations, that is, the result of using the mechanics was negative. And when personalised “bestsellers” were turned on, the average check increased by 5%.
2. Intuitive choice of strategy
As we said in the previous paragraph, you should not blindly follow existing algorithms. The solutions and mechanics you use should be tailored to the specific audience of a particular store, not its category. Usually in online stores, along with goods, they offer you to buy accessories for them. It is believed that if a person buys shoes, then they may well need a horn for shoes, which means that it must be shown on time – and the average bill should be increased. But you should not show other shoes – because there is a risk that the client will change his mind about buying these too. However, one research for online furniture and home goods retailers forces us to rethink traditional approaches. During the tests, one group of customers saw only a block of similar products, and the other – a block of similar and a block of related products. It turned out that when only related products were shown, all measurable indicators decreased: conversion, average check size, and profit. However, when both similar and related products were displayed on the page (in that order), the average check grew by almost 6.3% compared to the control group.
3. Lack of control
Even if an optimal solution has already been found for the tasks of an online store, it still needs to be constantly monitored, and the metrics must be monitored. Unfortunately, unforeseen changes that lead to a drop in profits can happen at any time. Make sure to have that in mind.
4. Same versions
Another fairly common mistake is using the same solutions for different types of devices. Just because one strategy is implemented and works successfully on desktops doesn’t mean it will perform the same on mobile. It is believed that people use different versions in different moods, exhibiting different types of buying behaviour. It is important to take these features into account both from the point of view of constructing recommendations and from the point of view of the design of the visual component.
5. Lack of critical thinking
Every marketing application is unique. You can’t blindly copy a working solution from another online store, or just take a standard one and hope that it works. This also works in the opposite direction: if a seemingly suitable solution is proposed, then it is still worth clarifying how it has already been implemented in other conditions. What if the mechanics look beautiful at the presentation, but the work does not bring the declared results and the resources for its integration will be wasted?
So, there is an opinion that algorithms based on image analysis should work well in the fashion segment: for example, when in the recommendations of related products, artificial intelligence determines the type of product and recommends the bottom to the top (for example, a skirt to a blouse). There is a study that compares recommendations based on user behaviour and recommendations based on image analysis. In most cases, testing did not reveal a significant difference between the algorithms, and in some cases, the “fashionable” algorithm showed a lower result. Thus, not everything that looks good in theory gives results in practice.
6. You treat every customer in the same way
For most stores, manually reviewing the details of each user and interacting with each, as is the case in B2B, is impractical. And this is where marketing comes to the fore. In B2B, a real person can manually analyse a post and decide how to create a relevant personal proposal. In most of the B2C world, this is not possible, given the number of customers you have to interact with.
For this reason, many online store marketers deliberately made a mistake and decided that since they cannot personally interact with each user and treat each contact as a real person, they will treat everyone the same. They send coupons to their entire contact list three times a week and call it email marketing. Instead of using marketing automation to solve problems by grouping contacts into separate groups that can receive more targeted offers, they show everyone exactly the same content on the site.
7. You Ignore user after purchase
Even intuitively, it’s easier to sell to people who once made a purchase from you, but the only thing many marketers do is send coupons to everyone who has bought before.
People get very excited the moment between placing an order and delivering it. Use the post-transaction time to communicate when the item will be delivered, submit relevant articles on how to use the product, and share your community on social media. Give your customer an opportunity to share the emotions of waiting with your community – capitalise on the excitement and attract the next users.
You can also keep in touch upon receipt of the product by sending out surveys that provide valuable insight into how your product is being perceived and nip potential problems at the root.
8. Consider that everyone has a universal reason for buying
A decade ago, Amazon was at the height of technology when they decided that if you buy the same product as someone else, you might be interested in other products that those users were buying. It was a victory. But eCommerce marketing has already outgrown this step.
Pay attention to your user-friendliness. Make sure to have a good quality web design. Take a closer look at your customers. Why are they using your product? If you’re selling camping gear, find out if your shopper is conquering the mountain alone or travelling with the family? The reason they buy from you is as important as what they buy. Rather than engaging based on what they buy, figure out a way (by analysing behaviour or by asking them) and use that in your marketing solutions.
9. Consistency versus user lifecycle
Another misconception in marketing automation is the idea that users should always move according to your plan. The traditional sequential logic of a marketer looks something like this:
User makes X → Send email and suggest doing Y.
Time passed, and the user did not do Y → Send an email and offer to do Y.
Another time passed, the user did not do Y → Send an email and offer to do Y.
Some time passed, the user still didn’t do Y → SEND SUPER CONVINCING LETTER AND PROPOSE TO MAKE Y.
Time passed, the user did not do Y → Stop sending emails to this user. He’s a bad lead and will never buy. Add to the master list and send the coupon three times a week, like everyone else.
This approach assumes that the buying cycle (which, despite its cyclical nature, is often depicted as a straight line) includes a set of actions that must be performed by the user. If he did action X, then the only possible option would be Y. What if I compare products and discover a new feature that I didn’t know about? I can go back to looking at options, and the type of content you send me should be consistent with these changes.
Instead, highlight a special segment that such users might fall into and warm them up with relevant content. The user life cycle is much more complex than a linear process. People can take a step back and that doesn’t mean they will NEVER buy. Consider this.
What is required for successful marketing? Before implementing a strategy, you should pay attention to similar implementation examples. The main focus is on the needs of the client: no one will tell what he needs, better than himself. Critical thinking is our everything: there is an individual solution for every case. Form hypotheses and test them, rather than relying on intuition and experience. First – mandatory testing, after implementation – control and ongoing support. Proven strategies can stop working, and that’s okay. The main thing is to react in time.