A Crystal Ball for Online Shopping

How search engine data is forecasting online retail sales for the 2015 holidays.

Wouldn’t it be wonderful if as online retailers we could predict what our next shopper from California is most likely to buy from our website – and when they are most likely to make this purchase?

Pullquote share icon. Share

Search and shop data has predictive value.

Search queries offer valuable insights into shopper intent. Shoppers search and then buy.

At NetElixir, using our proprietary tracking technology, we can track online shopper journeys from the time they click on a Google ad to when they make an initial purchase, as well as repeat purchases.

We are able to segregate the various touch points and websites involved (social media, email, affiliate websites, etc.) prior to a purchase.

Moreover, our technology assigns a time and geo stamp for each touchpoint, thereby enabling us to measure unique aspects of pre-purchase behavior.

For the past few years, our retail intelligence team has been diligently aggregating and analyzing millions of “paths to purchase” for seven retail categories, including: home improvement, apparel, home furnishings, gourmet food, consumer electronics, pet supplies and hobbies/gifting.

[Also on Longitudes: How to Plan for the Shopper of Tomorrow]

Pullquote share icon. Share

One out of five online holiday purchases will come from mobile devices this year.

Here are some of the insights we’ve found:

  • Women shopping online are more likely to buy dresses at 10 p.m. and home furnishings at 7:45 p.m.
  • Prior to buying a laptop online, a shopper clicks on or engages with more than seven websites – this takes an average of 16 days.
  • Shoppers from California take 13 percent less time to make a purchase compared to shoppers from Pennsylvania.

This in-depth study of shopper journeys helps us better understand the who, what, when, where and why of online shopping.

With more than half of online purchases influenced by search, we know this search-and-shop data has predictive value.

We can forecast the volume of future purchases at an aggregate level – at a greater than 70 percent accuracy rate – by analyzing historical search and shopping behavior.

The predictive value of search data also increases during the holiday season. We believe this happens because of the inherent “time pressure” online shoppers face to purchase gifts in time for Christmas.

Pullquote share icon. Share

Search engine data offers exceptional insights into consumer behavior.

In turn, this leads to a strengthening of “purchase-focused search intent.” For example, our research shows that search queries that include verbs like buy, get and purchase experience more than a 70 percent increase between Dec. 1 – 20.

We’ve set up a panel of 32 mid and large-size retail customers that we monitor and record through several key search metrics, as well as live e-commerce sales data (impressions, click-through rate, conversion rate, average order value, items per order, etc.) in six-hour intervals.

For predicting holiday e-commerce sales growth, we compare the first nine months of data for the current year versus three prior years. At the same time, we analyze the search query and online purchase trends for previous holiday seasons.

Our data analysis has produced several interesting insights:

  • The first “mini” online holiday sales spike occurs on the third Monday in October. To be more exact, it starts around 6 p.m. on Sunday and continues through Monday at midnight.
  • Seven percent of total online holiday sales come from last minute shoppers (Dec. 19 – 24).

Since 2011, based on our search-and-shop data, we’ve published a Search Marketing Holiday Calendar highlighting the top online shopping days for each year and the CPC (Cost-Per-Click) inflation.

Here’s the 2015 Search Marketing Holiday Calendar.

Click to Expand the 2015 Search Marketing Holiday Calendar.

Interestingly, the year over year, e-commerce sales growth projections based on our 32-retailer panel accurately aligns with the actual holiday e-commerce sales growth reported by the National Retail Federation in January.

This has happened three years in a row (2012-2014), leading us to believe that our sample – based on the retailers that comprise the panel – may reflect the overall U.S. online retail industry.

[Also on Longitudes: A Power Shift in Retail]

The projection for 2015

Based on our comparative year-over-year analysis for the first nine months of 2015 – including fairly strong back-to-school sales and continuing strength of e-commerce sales – we expect a year-over-year increase around 11.8 to 12.2 percent this holiday season for online sales. (This projection is entirely based on search-and-shop data and does not reflect economic factors like continued softness of the stock market and a volatile global economy).

In addition, we believe that more than 50 percent of all retail search queries and one out of five online holiday purchases will come from mobile devices this year.

Search Engine Advertising has been and still is a direct marketing channel that helps companies profitably acquire customers.

We believe that search engine data also offers exceptional insights into consumer behavior and has a strong predictive value.

Smart marketers can gain valuable insights into their online shopper behavior by carefully analyzing search-engine-led shopper journeys.

Moreover, with systematic effort, they can even succeed in predicting what the shopper is most likely to buy from their website. goldbrown2


Every morning, wake up to the blog that gives you the latest trends shaping tomorrow.

Udayan Bose is Founder and CEO of NetElixir.

Click the RSS icon to subscribe to future articles by this author. RSS Feed


We welcome the re-use, republication, and distribution of our content – just as long as you credit us. So we ask that you insert the following tagline when you use our content:

Reprinted with permission of Longitudes, the UPS blog devoted to the trends shaping the global economy.