How to Build an Alexa Skill for your Online Shop with Recolize in 15 Minutes?

Building an Alexa Skill is very easy if you combine it with Recolize personalized product recommendations. In the following I will describe the required steps and during the course of this blog post we will develop a sample skill that you can use as a starting point for your individual skill. The sample skill is available on our Github account.

We will create a so-called Custom Alexa Skill that will be able to simply answer the question “Alexa, open Recolize and ask what can you recommend to me today?” or “Alexa, open Recolize and ask what is the recommendation for today?“.
Therefore we need the Skill definition and additionally a AWS Lambda function that will trigger our Recolize recommendation API. The beauty of AWS Lambda is that you do not need to provision any servers but the so-called serverless framework will only be active if someone asks Alexa to trigger your skill. For more information on AWS Lambda functions see the official documentation.

We use the official Ask CLI commands to easily deploy both the skill itself as well as the Lambda function that will call the Recolize API.


  • First – if you haven’t done already – register for the Amazon Alexa Developer Portal in the upper right corner
  • Clone our sample Github repository and step into the directory.
  • Next step is to connect the Ask CLI with your Amazon Developer account:
ask init
  • Alexa Skills are triggered by the so-called “invocation name”, which means everytime you say “Alexa, open Recolize and ask…” the device knows which skill to activate. That’s why later on if you want to publish the skill under your name it is very important to put a lot of thoughts into a well-sounding, easy-to-remember and easy-to-pronounce invocation name.
    By default our sample skill uses the invocation name “Recolize” which you can change by editing the JSON files in the models/ directory.
  • Afterwards we deploy the Alexa Skill with the Ask CLI. The deployment steps takes care of all the required actions like creating the Lambda function as well as the Skill including the Intents.
ask deploy
  • The final step we need to do is adding some parameters to the AWS Lambda function which is done as environment variables. Therefore send us a message to request your Alexa Skill parameters and then go to the AWS Developer Console and switch to the AWS Lambda Function section and select the Lambda Function “whatCanYouRecommend” (which was automatically created by our Recolize Skill). Then scroll down a bit to the section “Environment Variables” and enter the parameters that we sent you there:
    AWS Lambda: add environment variables

See It in Action

Enough theory, now let’s get to practice and see how this simple skill looks like in action. The demo video below shows the sample application with the use of an Amazon Echo Dot device:

Some Final Advices

  • Testing of the Alexa Skill can either be done directly on the Alexa device or easily on your laptop with the excellent testing tool in the Alexa Developer Console
  • We have released this sample code under the GPL which enables you to use it in your own Alexa Skill projects
  • Later on if you want to publish your own implemented skill the official documentation for skill submission is a good lecture
  • Probably one of the first things you want to customize are the messages that Alexa replies. This can be done in the main JavaScript file lambda/custom/index.js. To only update the Lambda function after modification the following command is helpful after retrieving the ARN of the Lambda Function from the Developer Console:
ask lambda upload -f arn:aws:lambda:###FILL THIS WITH YOUR INDIVIDUAL ARN###:function:whatCanYouRecommend

That’s all it takes to build an innovative voice commerce skill!
Ok perhaps I am exaggerating a bit but as you can see this sample skill is very, very easy to build with Recolize and you can now leverage all the Recolize features like static product lists, extensive filtering and so on in Alexa.

What use cases come to your mind? I’m very excited to get to know.

ChrisHow to Build an Alexa Skill for your Online Shop with Recolize in 15 Minutes?
read more

3 Steps On the Way to the Perfect Title for Recommendations

Are you also bored by standard Amazon-like personalization titles like „You may also like“? Yes, of course, that’s a standard recommendation title everyone expects in online shops, but maybe a non-standard title is more surprising and will attract more customers?

In the following you’ll get a receipt to find out the perfect recommendation title for exactly your use case.

Step 1: Answer the following questions

  1. What is the average age of your customers?
  2. What target group are you focusing on?
  3. Are most of your customers female or male?
  4. In which context do you want to display your recommendations (on product pages, the cart page, in the blog)?
  5. Does the title also work on mobile devices?
  6. Is it possible to colorize your title?
  7. Are uppercase letters possible in your Corporate Identity (CI)?
  8. Do you want to display different titles for different target groups?
  9. Do your customers understand what personalization is or do they need some explanation?
  10. Do you want to focus on direct sales or visiting time increase?

Step 2: Analyze your answers

  1. The average age of your customers is very important for your recommendation titles. For young customers you can use a more personal language, for senior customers you have to be more formal. General titles work better for older customers, younger ones expect surprise, flexibility and freshness.
  2. Depending on your target group you can form your title in a very specific way. Despite the age in question 1, you also know from your experience how specific you have to get. It depends on the industry how exclusive your products are and if you have to tell the customers what’s contained in their recommendations. E.g. in a bike online shop a title like „Your accessories“ could work great. But the same title doesn’t work at all for fashion customers, because it’s not clear enough, which accessories are meant (bags, shoes, jewelry).
  3. Female or male, this is very important. Women expect smooth and nice recommendation titles. Men instead are more analyzing and want to know what’s contained and why. Think about which gender you’re addressing where. If you address both, then you need to use different recommendation titles. Then you have to answer these 10 questions for women and men separately.
  4. The context is defining what you are displaying. E.g. cross-sells only make sense on product pages and on the cart page. Make clear what you want to offer with the recommendation widgets in your title. „Others Also Like“ is not clear enough, a title like „Cross-Sells for Your Current Product“ is too technical and „Products With the Same Category Out of Your Personal Product Pool“ is too complex. If you mix up these three titles you could get something like „Matching Your Taste for Your Selection“. „Taste“ sounds personal, „matching“ sounds like you get the right recommendations and „selection“ refers to the current product. Great to understand for the current visitor.
  5. Keep in mind: nowadays most platforms have more than 50 % mobile traffic through tablets and smartphones. Your title must also work on mobile devices in length, style and content.
  6. Is it possible with your CI to colorize your titles? It’s nearly the same compared to a multivariate test with colored checkout buttons: think about giving your recommendations more visibility with colored titles.
  7. Uppercase letters are also a great way to highlight your recommendations. Again of course, it should fit your CI and it may not be to blatant.
  8. Think about displaying different titles for different target groups. Use e.g. English terms in other languages in the fashion area and use technical terms in stores that sell e.g. hardware.
  9. If your customers don’t know what personalization is or even worse, if they are afraid of titles like „We Know What You Want“, then don’t “frighten” them anymore. Then you have to make your titles more general and less personal.
  10. A very important question for your titles is, if you want to make direct sales with your recommendation widgets or if you e.g. want to increase the visiting time. Or do you want to get traffic back to your shops from e.g. product recommendations in a blog? Think about titles like „Your Personal Recommendations for Today“, so put more pressure on the customers to buy instantly within a time limitation, or a title like „Your Next Browsing Possibilities“ as an extremely opposite example.

Step 3: Combine and Decide

After working through the questions above you will have a cluster of answers and terms as base to decide on. Use the questions for every recommendation widget you want to add to your website. Take care for the context of the recommendation widget. The context is one of the most important criteria to match the interests of your visitors.

Also, after answering the 10 questions above, you won’t have the one and only answer, you will have a pool of titles you can work with and compare the performance of your recommendation widgets. Take your time to measure the different widgets and really compare their performances. Review your titles and the widgets constantly after defined time frames, because also seasonal changes can drastically impact the performance of your titles. These time frames depend on your business case, switching seasons etc. But you can take one review every month as a rule of thumb.

Good Examples

  • “Our Selection for You”: doubled personal, because your team chose the recommendations for the current customer
  • “Sale Products Picked for You”: specific, describes what’s contained and the customer knows that the recommendations are hand-picked for him
  • „Who bought this also bought“: this is a very well-known title – this can be positive or negative

Bad Examples

  • „You Might Also Like“: the customers won’t know what’s contained
  • „Personal Recommendations“: very general title and the word „personal“ may be to „spooky“ for the customer
  • „Who bought this also bought“: as mentioned in the good examples – very well known and therefore may not eye-catching enough for your customers. You have to test this one and check, if it works for your target group

The Impact of a Perfect Recommendation Title

You won’t believe it: the recommendation title is one of the first eye catchers for your recommendations and therefore the importance cannot be overstressed. The visitors will knowingly or unknowingly read this title and recognize if the contents match.

That’s also important: sometimes you just have to change the title to improve your recommendation’s performance by more than 100 %! It’s really unbelievable, but customers sometimes seem to not even recognize the recommendations with bad titles at all. By just improving the title with the rules above the performance has drastically improved. Now think about what’s possible when you not only change the title, but also configure your recommendations with a perfectly fitting recommendation rule?

With Recolize you can very easily switch your recommendation titles. Just create an endless number of rules, configure matching titles for specific urls and define fallbacks for e.g. unknown new visitors. You have countless possibilities.

Chris3 Steps On the Way to the Perfect Title for Recommendations
read more

15 % Conversion Uplift: Automated vs. Manual Cross-Sells

Many online shops are offering cross-sells. What you normally don’t recognize as a shop visitor: some of the cross-sells are manually created by the shop owner, others are automated. So what are the reasons for a shop owner to use manual or automated cross-sells?

What are Manual Cross-Sells?

Many ecommerce systems like e.g. Magento have built-in solutions to add cross-sells to product pages or the cart page. You then define manually for every product which other products should be cross-sold when the customer visits a product page.

When to Use Manual Cross-Sells?

Many of our customers – especially in the fashion area – tend to argue that they know best what their customers want to buy. Of course, the shop owner is an expert for his target group and therefore it absolutely makes sense, that he adds manual cross-sells for products where he e.g. knows, that customers in their retail stores always buy a specific shirt for a selected trouser.

But then, after we’ve talked to the customers and asked, if they’ve ever tested what happens when they offer perfectly matching other shirts for the taste of the current customer. Of course, this scenario has never been tested in retail stores because it is simply not possible. But in online shops you have a user profile, recurring customers and the possibility to complement your manual recommendations with automated ones.

Advantages of Manual Cross-Sells

  • Integrate knowledge from retail stores into online shops
  • Sell outs are possible for specific products
  • You as a shop owner know exactly which products appear as cross-sell for the currently selected ones

Disadvantages of Manual Cross-Sells

  • It is a huge effort to maintain the cross-sells
  • …and therefore very expensive
  • Visitors could not like the manual recommendations
  • You have to keep your recommendations manually up-to-date after e.g. season changes

What are Automated Cross-Sells?

Automated Cross-Sells can be used on every product page and on cart pages. Looking at the current product context, this kind of cross-sells automatically determine the perfect matching products for the currently selected one. This can be influenced by the conversion of the current products (“Others also bought”), but also by seasonal changes, the current stock availability, the categories of the product and much more.

Automated Cross-Sells combine personalization logics with product-based logics to get the best matching result out of the current visitor profile, combined with the knowledge of the shop owner regarding his product portfolio.

Advantages of Automated Cross-Sells

  • Automation for perfectly fitting products for the currently selected one
  • No effort to maintain the recommendations
  • Only the products currently in stock are suggested to customers
  • The conversion rate can be optimized by changing the rule set for the recommendations

Disadvantages of Automated Cross-Sells

  • You can’t exactly determine which recommendations will appear
  • Sell outs are only possible by “faking” the recommendations

Comparison of Manual and Automated Cross-Sells

A real comparison of manual and automated cross-sells is difficult, because except of the name the two kinds of recommendations don’t have any commonality. That’s also the reason, why we always have to explain, that “manual” cross-sells on product pages are actually only a manual way to add some kinds of recommendations on these pages if you have no automation. But: many of our customers still believe in the manual way and don’t want to switch completely to automated recommendations – at first. That’s why we recommend to use both: give your customers the chance to decide for you!

Add one manual widget next to an automated cross-sells widget and afterwards compare their performance. In our experience automated cross-sells nearly always convert better than manual ones. Which is no surprise in our opinion, because the automated cross-sells are not only suggesting product-based recommendations, but also use the knowledge about the current visitor to combine this with the product-based logics to then really suggest the best matching products for the currently selected one.

The results are amazing: e.g. our fashion customer HALLHUBER uses both. And compared to manual cross-sells the automated ones don’t only convert 15 % better, but also the revenues generated by customers using the automated cross-sells are amazing and nearly doubled compared to manual cross-sells. Therefore our customer is thinking to completely switch to automated cross-sells.

Conclusion: Don’t forgo the Revenue Uplift of Automated Cross-Sells

  1. Add automated cross-sells in addition to your manual ones and start with both: one widget with manual cross-sells and one with automated cross-sells.
  2. Compare the performance of your manual Cross-Sells to the automated cross-sells.
  3. Use an A/B testing tool to see what works better for you.
  4. For fashion customers: believe in the style of your customer. Of course you know which pieces should fit best for your customers, but maybe they have another opinion on this.
  5. If the results of your automated cross-sells are as amazing as for our customer HALLHUBER, then you can also decide to remove the manual cross-sell widget and completely trust in the automated way. Also offering two cross-sell widgets with slightly different configurations is a great conversion driver.

With Recolize you can easily add automated cross-sells to your product pages. Visit this article to see what’s possible.

Michael15 % Conversion Uplift: Automated vs. Manual Cross-Sells
read more

5 Reasons for Offering Automated Product Recommendations in an Online Shop Blog

Many online shops are also offering a blog to generate SEO traffic and inform their customers of interesting updates. Blogs are a great way to make an online shop feel more human and up-to-date when updated regularly. But as shop owner the main goal of a blog is of course to bring traffic back to the online shop and also convert blog visitors to online shop customers.

Therefore here are the 5 reasons why you should offer automated product recommendations in an online shop blog.

Reason 1: Make Your Blog Posts Sell

Most of the blog posts in supporting e-commerce blogs are talking about new products, changes in style, product seasons, a sale etc. After or while informing your customers, they should always have the possibility to buy something, which is missing in so many blogs. In static blogs the visitor has to navigate manually to the online shop and then search for the products he read about. Always offer these products directly in your blog posts and also other fitting products like e.g. personalized recommendations. This is easily possible with automated product recommendations without additional effort. Your blog visitors will thank you with a higher conversion rate.

Reason 2: SEO Traffic is Worthier Than SEA Traffic

An e-commerce blog mostly exists to improve generic SEO traffic for your online shop. The visitors coming through this sales channel were searching for exactly the content you are offering in your blog. Also, these visitors regularly stay much longer on your site and the chance to convert them is much higher than with dumb paid SEA traffic in your online shop. Because of these reasons you should offer product recommendations on your blog pages to make this worthy high quality traffic convert.

Reason 3: Avoid Potential Customers Leave Your Site

How are you behaving when reading a blog post? If interesting, you probably completely read the article, have a short look at suggested links below the article and if you came from a search engine, you use the browser back button to go back to the search results and visit the next interesting result. That’s where you lose your conversion chance as a shop owner!

Exactly at this point you have to interrupt the user by offering interesting and perfectly matching product recommendations. Make your own products more interesting than the next search results on Google. Only when you convince your visitors to visit your shop here, you have the chance to convert them. And of course product recommendations don‘t only make sense at the end of your posts – offer them also in between and give your posts more product context and flexibility.

Reason 4: Your Blog as Starting Point for a Perfect E-Commerce Customer Journey

Think about the traffic of your online shop. There are lot‘s of starting points for your onsite customer journey:

  • Paid SEA visits
  • Returning customers
  • Newsletter clicks
  • Generic SEO traffic

Especially the last one, generic SEO traffic, is the one where your potential customers at first get in touch with your blog. So the starting point for your online shop‘s customer journey then suddenly is your blog! Then you have to ask yourself: „Ok, now I have lot‘s of generic SEO traffic, but how can I make them visit my shop?“. And one of the answers is: use the recommendation engine which is already used in your online shop. Then you have a win-win situation: on the one hand you improve the performance of the recommendations also with your blog traffic and on the other you use the intelligent recommendations generated by your shop‘s traffic in your blog.

Reason 5: Automated Product Recommendations are Always Up-to-date

You probably know this szenario: you are constantly creating great blog posts with static links back to your shop products. But with a season change all these products are gone. Or your products are out of stock. This can be really bad for your customer journey and also your SEO rankings. With automated product recommendations instead of static links you always only offer currently available products. You can also fall back to personalized recommendations if e.g. the season is changing. So you avoid additional effort to update old blog posts and dead links and therefore disappointed blog visitors will never be a topic on your todo list again.

Automate Your Blog Product Recommendations

With the 5 reasons above you‘ve learned why you should no longer pass on automated product recommendations in your online shop blog. You can completely automate the recommenced products, guide the traffic into your shop catalog and improve your conversion rate with this traffic.

If you are searching for a personalization solution offering this, try Recolize. Our unique cross-domain feature offers the possibility to integrate your product recommendations e.g. into your WordPress blog with 2 clicks. Furthermore you have the possibility to integrate your recommendations also on all your other domains like e.g. on your company website. Visit the reference page of our customer Resch&Frisch who does exactly this on his TYPO3 company website.

Michael5 Reasons for Offering Automated Product Recommendations in an Online Shop Blog
read more

6 Reasons for Automated Recommendations on 404 Pages

The power of 404 pages is often underestimated. “Not found” pages are of course not the #1 conversion generation pages, but you should really invest some thoughts in creating reasonable 404 pages instead of displaying a standard “Ooops, we’re sorry.”. Help your lost visitors on your 404 page. Here are the 6 reasons why you should do that and offer automated recommendations on these pages.

Which Traffic can Result on 404 Pages?

There are lot’s of possibilities why your visitors are landing on a “not found” page:

  • Old links from external sites
  • Missing redirects for old pages
  • Old SEA links e.g. in Google AdWords
  • Wrong internal links
  • Deployments of new versions
  • Partial downtimes of your website

Whatever reason, it is lot’s of effort to solve all of them and avoid 404 visits completely. That’s why you should really invest some time in optimizing the 404 page, because it will probably be displayed to a significant part of visitors.

Reason 1: 404 Pages Often Have Lot’s of Traffic

Especially when you have an online shop and generate paid traffic e.g. through Google AdWords or Bing Ads, it is always possible, that you buy traffic for product pages which are already offline because the products are e.g. out of stock. The product feeds for the mentioned paid services are normally updated only once a day, but many online shops are updating their product catalog much more frequently. This can result in 404 pages for at least some hours a day for very expensive traffic.

Reason 2: Help Your Lost Customers

When visiting a “not found” page, most of your customers will be irritated. They intended to visit a specific page, but it is gone. Therefore it is very important to help your lost customers not only by displaying them that they are lost. This is what most 404 pages do and your customers will be even more irritated and the chance that they will leave your website is very high. Therefore you should help your customers by saying “Sorry” and propose next steps by displaying relevant recommendations.

Reason 3: Don’t Interrupt the Browsing Experience

404 pages are normally visited as first page or somewhere in the middle of the customer journey of your website visitors. This customer journey must be as smooth as possible to convince a potential customer of buying or a visitor to stay even longer. Therefore “not found” pages should not look like “sorry pages”, but also contain relevant content for the current visitor. The customer has to think: “Oh no, the product I’d like to visit is gone, but hey, I already see great matching similar ones. Let’s move on and buy something different”.

Reason 4: Help Your Broken Links

Broken links are nearly unavoidable. In online shops you will always have products that a suddenly out of stock and in blogs there will always be external broken links to your website. Of course you should repair all these broken links with redirects, but you should also care for the time in between fixing the causing issue and when the issue started. A great 404 page is one perfect possibility to mitigate the issue as broken links will always occur.

Reason 5: Convert Traffic on 404 Pages

Why shouldn’t 404 pages be conversion optimization pages? As written above, there are several reasons why 404 pages can become lot’s of traffic. So, you should conversion optimize it like e.g. your home page or any other landing page. Add automated and relevant recommendations, beautiful teasers and images and e.g. coupon codes. This will convince your customers to continue and improve the conversion rate of formerly lost visitors. A great chance to increase your overall performance and to not lose all the traffic landing on 404 pages.

Reason 6: Automation vs. Manual Updates

A great reason for recommendations on 404 pages is the automation possibility. It is easily possible to keep your 404 pages up-to-date like your home page, but normally you don’t have additional time to also update your 404 pages manually. Therefore automated recommendations or product recommendations are perfect for keeping “not found” pages current and investing no time in doing so. The recommendations update and improve automatically, and your customers always get displayed perfectly fitting proposals.

Summary: Take action!

Invest some time in your “not found” pages. The integration of automated recommendations to cross-link your customers back to your website is a great possibility to do so. This can be done once and afterwards all your different 404 pages display relevant content to your visitors – no matter where they came from.

With Recolize you can easily integrate several recommendation widgets on 404 pages. Many of our customers are displaying two widgets: one with personalized recommendations for the current visitor and the second containing last viewed products to improve the browsing  experience of the customer.

Michael6 Reasons for Automated Recommendations on 404 Pages
read more

HALLHUBER wins 2. Place at the Shop Usability Award

Our customer HALLHUBER relaunched its online shop in February and moved from Magento 1 to Magento 2. With this switch a beautiful new design and usability concept was implemented.

Directly after the relaunch HALLHUBER participated in the german Shop Usability Award and won the 2nd place. We are very proud to be part of this award-winning new shop and we’ve implemented a completely individual design for the online shop based on their amazing new concept. In the screenshots above and blow you can see the Recolize integration on the home and on a product page.

Recolize on a Product Page of HALLHUBERCongratulations to HALLHUBER and thank you for your support.

What Does Recolize Do for HALLHUBER?

Recolize optimizes the onsite traffic of the online shop. With our recommendation widgets integrated into every important shop page, the shop increases its conversion rate and average order value drastically. Recolize is integrated into the following pages:

  • Home
  • Product pages
  • Cart
  • Wishlist
  • Mobile cart layer
  • Search results
  • Checkout success
  • 404 pages

On the product page, the search result page and on 404 pages our customer decided to use more than one recommendation widget to reach a perfect customer journey through their online shop. Offering the possibility of a browsing history with last viewed products is also a feature of Recolize they are using on these pages, just as automated cross-sells next to manual cross sells on product pages.

What are the Results of the Recolize Integration?

After some months with the updated Recolize integration in the new online shop, the KPIs of the integration are amazing. Over 25 % of all the customers are actively clicking on a recommended product by Recolize before buying something. This number increased around 10 % after the relaunch thanks to our flexible integration into the new design. Furthermore 7 Recolize domains for all countries and languages are used and in one month now already over 4.5 million recommendation widgets are offered to the customers.
In the updated HALLHUBER case study you can now also find all the great results Recolize generates for HALLHUBER.

MichaelHALLHUBER wins 2. Place at the Shop Usability Award
read more

3 Ways to Integrate Cross-Sells on Product Pages

Lot’s of customers ask us how to integrate cross-sells into their product pages. Of course, the definition of cross-sells differs from customer to customer. Some are understanding manually selected fitting products to the currently selected one as cross-sells. Others want to automate. This article describes the automation: 3 ways to integrate automatic cross-sells into your online shop’s product pages.

What are Cross-Sells?

Cross-sells are defined in many different ways. Assuming that we are talking about an online shop, then cross-sells are offered on product pages. Cross-sells then are product-based recommendations and fit to the currently selected product. In comparison to real personalized recommendations, cross-sells have another context and only really work on product pages, because they need the currently selected product to function properly.

Way 1: Display Products From the Same Category

In many shops you get recommendations for products from the same category. This can make sense for shops with low-priced products like e.g. fashion stores. If you think of high-priced products like e.g. in bike online shops, then products from the same category don’t make much sense. Cross-sells are per definition additional products your customers should buy on top of the currently selected product. If you really want to display products from the same category, then please think of the pricing of the currently selected product and if it really makes sense to offer similar products within the same pricing range. If you still decide to display products from the same category for high-priced products, then you just offer one more way to improve the customer browsing journey through your shop, which can also make sense in specific cases.
To sum it up: cross-sells containing products from the same category make sense for low-priced products or product variations (e.g. sizes, colors). For high-priced products you have to think about different logics.

Way 2: Display Accessories for the Current Product

Proposing accessories as cross-sells is a great possibility to display matching products which don’t conflict with the currently selected products and also don’t interrupt the current user journey. To do this, you explicitly have to exclude the categories of the currently selected products to avoid similar products.
Try the following to find out how to configure your accessories rule:

  • Answer this question: which products do you think your customers would buy on top of the currently selected product?
  • Analyze your quotes, have a look at what your customers are already doing and narrow this down to a more specific rule.
  • Don’t only create one general rule for all cases. Cluster your products in categories or price ranges and then create separate rules for these clusters.

Way 3: Display Lower Priced Products

This is not only a great possibility for your product pages, but also for an online shop’s cart page. On product pages you can offer low-priced products for a high-priced selected product. Or just display low-priced products in your cart without any context. An example can be a bike bell for a mountain bike. So you can easily increase the average order value and push sales for cash and carry products. This can be compared with the checkout in a supermarket: these tiny little chocolates and we always think: “Ok, just one for the way home”. Try to reach this experience in your online shop.

Does Recolize Support Automatic Cross-Sells?

In the Recolize Tool you have lot’s of data pools available. One is “Cross-Sells” which is exactly what you need for offering automated cross-sells. In combination with filtering the attributes from your product feed you have a very powerful marketing tool to improve your conversion rate.

The Logic Behind Recolize Cross-Sells

Recolize cross-sell recommendations are not static at all, but depend on the automatically generated product recommendations based on your visitor’s behavior. For all 3 ways described above, this data pool should be the base of every cross-sell rule.

Michael3 Ways to Integrate Cross-Sells on Product Pages
read more

Recolize wishes Merry Christmas

As both last years (see our Christmas post in 2015 and Christmas post in 2016) since our foundation we wish you – our customers, partners and Recolize fans – wholeheartedly a Merry Christmas.

We can look back proudly on several successfully reached milestones:
We were able to quintuple our customer base and therewith generate and display millions of recommendations for our customers.
Our focus on simplicity and transparency fully pays off for our customers: With an improved Customer Intelligence Module we are able to pinpoint and guide you even more targeted to possible problems and improvement hints. You do not have to be a personalization expert, Recolize is always by your side as the smart assistent.

Additionally we are also proud to be an official PrestaShop partner now and therewith extend our ecosystem of shop software by a well-known solution. As accustomed with our Magento, Shopify and  WordPress plugins, the installation is done with one click and the recommendations are instantly shown in your PrestaShop – that’s how easy personalization can be in 2017!

Furthermore we have developed our Recolize Extension for Magento 2, so that you can also use the latest version of this popular E-Commerce software with Recolize.

Of course we are also happy for some reflective days now to start full of beans and with fresh new ideas together with you into 2018.
We’re looking forward to very exciting times!

Your Recolize team,
Anja, Michael und Chris

ChrisRecolize wishes Merry Christmas
read more

Recolize Becomes PrestaShop Premium Partner

PrestaShop is one of the world’s most popular E-Commerce platforms with a growing international community. Currently, it is used by more than 270,000 eCommerce retailers. With just one click, all of them can now offer our personalized product recommendations to their customers. The installation of the Recolize module takes place directly in the PrestaShop Back Office.

Easy Installation of Recolize Module in PrestaShop

Recolize offers an unlimited number of recommendation carousels with personalized products, cross-sells, up-sells, best-sellers or recently viewed products that can be integrated on every web page. In addition, our PrestaShop retailers can personalize their product recommendations across all domains, for example, blog, E-Commerce shop and company websites. Thanks to our smart assistant, all data is collected in real time. Therefore, anyone can quickly understand how Recolize personalization help to achieve the goals and maximize the revenue.

We support all PrestaShop versions 1.5, 1.6 and 1.7 in many languages (German, English, Spanish, French, Polish, Italian).

Start right now! 
And install our Recolize module directly from the PrestaShop Addons-Page.

About PrestaShop
PrestaShop was founded in 2007 with the aim of providing free E-Commerce software using open-source technology. PrestaShop features in the Inc. 5000 2016 list, which includes rapidly growing private companies in Europe. PrestaShop won the People’s Choice CMS Award for best SME e-commerce solution. PrestaShop has offices in the USA and France and is financed by XAnge Private Equity.

MichaelRecolize Becomes PrestaShop Premium Partner
read more

How to Migrate your Extension to Magento 2?

In this blog post I will show you a sample how to port your existing Magento 1 extension to the brand-new Magento 2 based on our Recolize Recommendation Engine extension for Magento 2.

While Magento 2 was publically released in December 2015 more and more extension providers are migrating their extension to the new platform. As the whole architecture of Magento 2 is different than in the first version the steps required for making the extension compatible with the new Magento version differ a lot from extension to extension depending on the complexity and used parts of the Magento system.

Most of the steps described here are taken from these sources:

(Generally for sources on the web it is important to look at the creation date as a lot has changed since the first official Magento 2 beta version.)

0. General hints

Magento 2 now has a nice command line tool bin/magento (comparable to the well-known n98-magerun in Magento 1) which offers lots of cool commands.

Assure to set the Magento 2 Developer mode as described here with

bin/magento deploy:mode:set developer

as well as to disable the Full Page Cache (FPC). This assures that e.g. static files are not cached which makes developing extensions a lot easier.

Magento 2 now ships with XSD files to validate your XML files. This can be integrated directly into PHPStorm which is a very handy feature. Just execute the following command (see here):

bin/magento dev:urn-catalog:generate .idea/misc.xml

Please note that there are also automatic converting solutions like the Unirgy Magento 2 Converter but we will do that manually for sample reasons.


The required steps for migrating the extension are:

  1. Create a skeleton Magento 2 extension
  2. Migrate system settings, localizations
  3. Migrate extension template stuff
  4. Migrate extension product feed generation logic

1. Create the skeleton Magento 2 extension

As the whole architecture and therewith also the file structure is different from Magento 1 it is advisable to create a completely new extension package.
The basic file structure looks like as follows:

After creating the basic file structure you can execute

bin/magento module:enable && bin/magento setup:upgrade

and the extension is now installed which you can easily verify by visiting the Admin panel Stores > Configuration > Advanced > Advanced

2. Migrating system settings

This task means mainly converting the old system.xml file into the new Magento 2 XSD format. As you can see from the picture below that shows the comparison between old and new format the changes can easily be understood:

For localization files it’s even easier as you only have to move them to the i18n folder.

3. Migrating extension logic

Next step is to migrate the existing logic from our Magento 1 extension.
Basically our Recolize Recommendation Engine extension does the following to major things:

  1. Add the layout handles for integrating our JavaScript snippet and the Recolize JavaScript parameters on special pages like cart, product detail page, etc.
  2. Update the user information in the cookie on login/logout.
  3. Periodically generate the product feed (see chapter 4).

As these migration tasks are very specific to our Recolize Magento extension it’s up to the reader to check out the necessary steps and differences with Magento 1.

4. Migrate extension product feed generation

In Magento 1 we relied on the Magento default Dataflow extension to generate our product export feed because there many flaws have already been implemented and made some customizations (see my earlier post about Dataflow customizations).

As Magento 2 only ships with the ImportExport module and that functionality is very limited we decided to implement our own feed generation methods. But that’s part of another blog post..

And basically that’s it! As you can see for simple extensions like ours that do not go deep into the Magento internals it is mostly re-writing some XML stuff into the new format and adapting to the new dependency injection logic.

Now it’s time to try out yourself. See our Recolize Magento 2 extension for a sample.

ChrisHow to Migrate your Extension to Magento 2?
read more