Blog

Anyone who has ever shopped online has been in this situation at least once (in fact, more than once). Something that you never intended to buy suddenly appeared in the “recommended products”. How did it end up in your recommendations? It’s just that “smart” Big Data has once again decided everything for you, for the marketer of this store, and for everyone in general.

A typical black box

With an unpredictable result. It’s impossible to calculate how much this harms your sales, since recommendations are calculated individually for each visitor. And no one knows what your customers actually lebanon phone number data see in the “recommended”. However, there is a problem.

Everyone already knows about personalization, and everyone knows that it is good for sales and customer loyalty. But there is a caveat here – this only applies to the personalization that works correctly.

What do I mean by “right/wrong”? Any online store is about industry specifics. This means that the behavior of visitors is determined by very individual patterns that are valid only for this market. Or even for this site.

For example, look at any children’s store

There is such an obvious characteristic of the product as the age of the child. This is, one might say, a determining factor, it sets the behavior of the parent-buyer. If your child is three months old, you do not look at metal construction sets, this is logical. Accordingly, personalization in the store should work in such a way as to take this factor into account.

But on the other hand, there are manufacturers of product recommendation engines. You connect one to the store, and you get personalization. They, in the vast majority, have never heard of how to make a balance sheet in detail industry specifics. The key task of such a business is to build the most universal product so that it works more or less normally with children’s goods, and with auto parts, and with cosmetics.

As a result, something like this can happen to the store:

Why toys for children aged 4-5 ended up in related products – I’m sure data on no one knows. To be fair, it should be noted that if you scroll through the product description, you will find a block called “similar products” at the very bottom of the page. And there is something that looks like the truth:

You can talk endlessly about the effectiveness of product recommendations, but it’s all useless if they are located in the wrong place on the site.

Check your store with our checklist. Examples below.

Where to place

In principle, recommendations can be placed on almost all pages where products are presented. More often, one or two such blocks are enough for one page.

The logic of each such block should correspond to the stage of purchase at which the visitor is currently. For example, if a person is on the search results page of the site, he is at the stage of active selection, comparing products, is going to buy something now or later. If on the product page, he is close to buying, now you need to offer him the most similar products-analogues in a suitable price range, to make it easier for him to make a decision.

As for the place on the page, there are two rules:

The buyer should not scroll for a long time to get to the recommendations laos phone number data from the main area of ​​interest (for example, a block with a photo of the product and its characteristics)

Recommendations should be in the main content area (and, for example, not in the footer and sidebar).
Personalized product recommendations

Increase the purchase receipt, make cross-sells. 8 ready-made sales scenarios and unlimited customization options. Launch A/B tests with one button.

Home, category and product page, cart, search results, 404 – the main pages where you can most often see product recommendations. More about them.

Home page

If the visitor starts their journey from the main page, then most likely they do not have a specific goal. Therefore, it is logical to tell about the latest events and introduce the product range on the main page.

This is usually implemented using blocks – “Popular products”, “You may like this”, “Product of the day” and “Buying right now”. If you do everything right, there is a chance to increase conversion by 10-20%.

“Popular Products” Block
The most suitable products for the client among those popular on the site.

This is the most “general” selection, which makes sense to show when we don’t non-standard channels in performance marketing know anything about a person. We don’t yet know their desires and preferences, so we just display what they buy most often.

Where to place: In the main content area. An important point – we place promotions data on and new products at the top, since it is important for us to draw special attention to them and provoke a person to click and go to the catalog. Recommendations – right below them.

We recently talked about how voice search is changing not only the typical queries on the Internet, but also people’s behavior in general. Retailers will have to make an effort and optimize their site for voice search before it completely displaces text search – otherwise, there is a risk of falling behind competitors by a couple of corpora.

Voice search optimization

or VSO, is becoming increasingly important. According to researchers, by next year it will account for 50% of all search queries.

As neural networks learn and technologies develop, voice search kuwait phone number data becomes more intuitive and begins to capture the

nuances and context behind queries. Let’s look at the key factors to pay attention to when optimizing your site for voice search.

Natural language when creating content
Remember the set of key phrases that copywriters had to squeeze into product descriptions? The resulting texts were little like human language, but now everything is changing. When people use voice search, they don’t limit themselves to stubs of phrases, but speak normally — as if they were talking to an old friend.

That’s why it’s time to include everyday phrases in your content — this is to attract voice search engines to the content of your site. Long keywords in titles and descriptions are a great way to get a higher ranking in search results. But this is not the only place where natural language is important.

The fact is that due to incorrect pronunciation

voice search is not as accurate as text queries. And the more keys you scatter throughout the site (and not only), the higher the likelihood that the user will notice you. Where to place them:

Although voice search may not lead a potential buyer straight to the capricious google product page, they will learn about your brand and follow the links to the site.

Optimizing content for the needs of visitors
A loyalty system, brand benefits, payment and delivery terms, blog articles — all this attracts your potential customers. Every retailer needs to anticipate what questions their target audience might have while searching for a product, and what content is important to them.

1. Voice search is a complex tool, but every time your content is in data on the spotlight, your site becomes more valuable to search engines. Knowing what users will be looking for and answering their questions in advance means driving traffic.

In the US, more and more online store buyers are switching to the subscription model: last year alone, their number grew by 15%. Russian e-commerce is just getting started, but no one doubts that the trend, as always, will come to us a little later. So let’s get ready – get acquainted with subscriptions.

The growing popularity

of subscriptions is a natural process, and it is easy to explain. First of all, it is convenient for customers: instead of buying the same thing once a month, they simply pay money and receive the product directly korea mobile phone number data to their home. Those who have already

appreciated the benefits of subscriptions do not limit themselves to one: more than half of online store buyers have at least two.

For retailers, the benefits of subscriptions are obvious:

Savings on attracting new customers. The standard model, which is based on conversion, means that you have to spend a lot of effort and money on attracting customers. First, you turn a visitor into a buyer, and then work hard again so that he becomes a regular. The subscription model means that your customer immediately becomes a regular customer, and provides a stable income.

More accurate forecasts. To forecast income, traditional businesses have to rely on a number of assumptions and use a lot of analytics. The subscription model allows you to estimate income much more easily: you take the total number of subscribers, subtract the cancellations, and multiply by the subscription cost.

Optimized inventory management

v

Imagine: you bought a product in advance, but it does not sell as well as expected. Or, on the contrary, a product suddenly goes viral, and its stock is running out, and you still have a thousand omnichannel customer experience strategy for saaS enterprises customers who want to order it. With a subscription, these problems disappear: you know exactly how many subscribers you have and how much product you need – you can plan your purchases more accurately and virtually eliminate leftovers.

And the best thing is that today, retailers have everything they need to implement a subscription model. They have many successful examples of companies and opportunities to use their experience in their niche.

Online books
In general, subscription-based models are ideal for two types data on of businesses: those who sell content, and regular services. We will return to the latter soon, but for now let’s talk about content in ecommerce – we are talking, of course, about the online book market.

Quite a lot has been written about the need to do internal store search. Mostly they talk about various usability improvements, like “make the search field noticeable” or “get rid of zero results”. We even once wrote a comprehensive collection of requirements for modern e-commerce search — we listed what it should be able to do and what it should look like. The text is completely relevant even a year later: a huge layer of online stores still simply ignore the search. Although, it would seem, there is hot traffic, interested visitors.

In general, everything is simple here

if the search engine collects statistics on search queries, this data can india mobile number data tell the marketer what to do. Using the example of REES46 search, I will show what exactly can be learned from the search and how to respond to it

Fantastic queries and where they live
What you should pay attention to: popular queries, queries without results, queries with maximum revenue and popular queries without revenue.

Popular queries (and popular ones without revenue)
The first thing to track is popular queries. In fact, this is a way to measure the demand of your visitors using the search tool.

What period do they

refer to: a week, a month, six months? This way you can track the stability of demand for a particular product. If a query stays at the top for a long time, it makes sense to add it to the purchasing plan. I won’t judge what period will be sufficient, industry specifics rule here. For example, in bookstores, the popularity of a query is very the what, why, and how of a/b testing in digital marketing easily checked by the general hype around the book in the media. And the hype itself lasts for a couple of months at most, after which it declines – so it makes sense to respond to trends quite quickly.

The first line in popular queries is a query for books by the writer Alisa Ganieva. There is a surge in demand for her books now, this is a signal to purchase the product.

What conversion do these queries give: zero or non-zero. If the conversion rate is above zero, it means that the goods are bought directly through the search. This means that visitors coming to the site with a need data on for such a product are the hottest of the entire top. It is worth bringing products with high-conversion queries to the main page – to further increase sales.