The Marketing Technologist.

We talk about analytics, code, data science and everything related to marketing technology. Backed by the tech geeks of Greenhouse.

Multichannel problem (partly) solved with postal codes

Postal codes play an important role in today's e-commerce. They are used to get online orders delivered to your home, food boxes sent to the right address and to get postmen to their destination as fast as possible. So, it's obvious that the postal code is essential in the logistic process, but what more can we do with it?

Is it possible to use the postal code for other, even smarter purposes?

One of our clients, a furniture retailer, has almost one hundred physical stores across the country and runs a very successful web shop. As with a lot of retail companies, their omnichannel approach is a present-day topic for them. Whether the customer is shopping online from a desktop or mobile device, or offline in a physical store, they deserve a seamless experience. How can we serve the customer across all points of sale, and how can we apply our online marketing activities to support this multichannel strategy? And, more importantly (at least for us), how can we measure the offline effects of our online efforts?

Project 1234 AB

Enter 'Project 1234 AB' (1234 AB is the format for postal codes in The Netherlands): an experiment in using postal codes to solve the multichannel problem of not knowing what your online customers are doing offline and vice versa. So, how do we get those postal codes? On the website of our client, there are three places where the customer is asked to leave their postal code:

  • The product detail page: by entering the postal code, the user can check the stock availability of a product in the nearest store.
  • The Store Locator: enter your postal code to find the nearest store.
  • With every order, to make sure the products are be delivered to the correct address.

By registering the postal codes on one of the three mentioned moments in Google Analytics, we can match them to offline sales. Because the retailer also asks offline shoppers for their postal codes with every store purchase.

Of course, the postal code itself is not unique to a single customer: you'd need the house number too. Unfortunately, the house number is not collected on the website. Based on our analysis, it turns out that the total of sales in most postal codes is tiny. Thus, only having the zip code (without the house number) is sufficient to get good insights.

It’s important to realise the postal code match doesn’t give us an inclusive overview in multichannel behavior as a whole. There are users that won’t fill in their postal code, simply because they already know where the closest store is. Also, if a user checks his postal code at home and decides to make a purchase at a store near his office, it’s impossible to match the search to the specific purchase. Then there's always a percentage of offline customers not willing to give their postal code.

Insights

Despite these limitations, matching the postal codes has lead to the following insights:

  1. Research online purchase offline (ROPO)
    We all know there's a connection between visits to a website and purchases in a physical store. Matching postal codes made it possible to proof the correlation for our client. Despite the mentioned limitations, the ROPO (or multi-channel) effect was far bigger than expected. A significant part of the turnover by offline purchases could be traced back to a customer's orientation on the website.

  2. Revaluation of the media mix
    Now, with the matched postal code data in place, we can measure our online marketing campaign's success by the number of sales in the webshop, but also by the sales in the offline stores. Because we can now optimise on the total turnover, we see a shift in budgets and a change in the online media mix. Media that hadn't been cost-effective before, have been analysed again by allocating the total turnover to different sources. This has lead to a revaluation of the online media mix.

  3. Retargeting strategy
    The postal code data gave us another interesting insight. By matching the date that the postal codes were collected on the website to the time of purchase in the psychical store, we were able to see how much time generally passed between these two moments. This, of course, is of exceptional value to your retargeting strategy. It gives you leads on how long you should retarget and what message is most effective in this effort.

Multichannel impulse

This first analyses on postal code level was a simple experiment, but the insights were outstanding. Reason enough to start automating the process and structurally analysing the data. 'Project 1234 AB' has given multichannel thinking a big impulse. At first, the postal code was only needed to get the ordered products to the customer. Nowadays, the four numbers and two letters are important data that are needed to better understand and service the customers, through every channel.

Cover image: Alastair Thompson, licence: CC BY-NC-ND