Compete Blog: It’s 10 p.m. Do You Know Where Your Shoppers Are?
Through a special arrangement, presented here for discussion
is a summary of a current article from the Compete Blog. Compete Inc. is a
web analytics company that focuses on understanding how consumers use the internet.
One
of the many reasons people love online shopping is the freedom to place orders
anytime, anywhere. We’ve all bought a pair of shoes at 3 a.m.
once, right? (Maybe that’s just me.)
I was curious to see how many people
actually are shopping in the wee hours of the night, and what times of day
are most popular for e-commerce. I broke down the day into one-hour segments,
and looked at average daily visitation to several major retailers and categories
during each hour. While the overall trends were similar across the board, a
few interesting patterns emerged.
Visitation to most retail categories peaks
in the evening hours, around 8-10 p.m. There is a steady increase throughout
the day (from about 9 a.m. on), and it drops off around 11 p.m. The lowest
levels occur in the early morning hours — between
3 and 5 a.m. Though traffic drops dramatically, there are still diehard shoppers
at this time of day, proving that the world of online shopping never sleeps.
Categories behave differently. For example, Sporting Goods retailers see a concentrated
peak around 8 p.m., after fairly low levels throughout the day. Home Improvement
sites peak in the morning and remain steady through the afternoon, with an early
drop-off at night.
I also looked at four major online retailers — Amazon, Overstock,
Target, and Walmart –and compared their shopping patterns throughout the day.
While the lines trend similarly, Amazon and Overstock skew slightly later,
with peaks at 8 p.m. Target and Walmart see visitation a bit earlier, with
peaks at 7 p.m., and more dramatic drop-offs at night.
Finally, I wanted to
see if certain demographics had different visitation patterns. I took the Apparel
category and broke out hourly visitation by male and female groups. Both had
peaks in the evening around 8 p.m., however, women had more constant visitation
throughout the day, as well as a peak in the morning.
Using such insights, apparel
retailers could target women during these peak times, or work to draw more
men to their sites throughout the day. But knowing when their prime customers
are online, retailers could potentially target their online buyers more easily
and potentially bring more traffic to their site — much
like many try to do with their in-store buyers during peak times. In the world
of email and social media bombardment, retailers have to work harder than ever
to stand out from the pack.
Discussion Questions
Discussion Questions: What insights can be derived from knowing daily shopping patterns of online buyers? How does it complement online marketing based on learned purchasing behavior? What is the best way to target consumers for a specific time of day?
Digital is an action environment so brand building, path to purchase, and shopper insights all come together, hence, the phrase digital path to purchase. I use Compete data in this way on behalf of a client and it offers a major contribution to actionable insights, sense, and response, so to speak, and it will help to change beliefs about how marketing can create growth.
Let’s see. Most people don’t shop online when most people are sleeping. Hmmm….
And people do more online shopping when they have freer access to computers. Hmmm….
And, it makes sense to try to target potential shoppers when they have the greatest potential to be shopping. Hmmm….
Hmmm….
Knowing visit patterns is important for operations, such as servers, but shouldn’t be over-valued in marketing. Marketers gain efficiencies by targeting peak hours, but they won’t necessarily grow sales when the challenge is reach. Two comments here:
First, contextual and behavioral targeting already address desired shopper tendencies and mindsets. For the large retailers with large sales goals mentioned here, further slicing by daypart has modest impact.
Second, advertising to brick and mortar shoppers during peak periods is not quite the right analogie because of vast difference in numbers.
But when large manufacturers and retailers choose display advertising, the goal is not simply immediate response, but also building associations that lead to future purchases.
Studies showing “banner blindness” lead us to value immediate response. But ads viewed unconsciously still create long term memories. They are processed similarly to overhearing your name in a nearby conversation. Immediate metrics like hourly periods miss the full story:
1. The peaks aren’t that high. The Compete chart doesn’t show peaks 5-10X the average, but rather 2-3X at best. Ignoring the valleys leaves shoppers unaddressed.
2. Time is displaced. Shoppers research before they buy. Further, Target, Walmart, and other multi-channel retailers use site-to-store programs where browsing and buying is disconnected.
3. Last click is a problem. When controlling for final attribution, display ads have more purchase influence than initially measured.
4. Manufacturers’ ads influence sales elsewhere. Amazon, Overstock, Target, and Walmart attract many shoppers who buy on other e-commerce sites.
I love the charts–great information!
Retailers are now in a position to offer time-sensitive promotions. I mean really time-sensitive: Discounts for people ordering online during a commercial break from The Closer. Or say, The 3 a.m. Insomniac’s Special. Probably best would be incentives to shop one site over one’s competitor during the “rush hour.”
Great stuff Blogger–Keep it coming.
I love the insight and it stimulates thoughts as to how marketers can fine tune their messaging to customers.
The difference between how brick and mortar and online retailers can effectively use information about footfall by day-part is related to capacity and resourcing constraints.
Brick and mortar stores have logical motivation to steer shoppers to the store at off-peak days and times. There are obvious benefits to be derived.
The same cannot be said for online retail. The benefit would come more through highlighting specific merchandise, even negotiating advertising rates with suppliers based on the most heavily visited times of day by potential purchasers.
Thought provoking data. Hope this conversation continues.
I hate to be cruel, but I have to go with Ryan here: could there possibly be a more obvious set of observations than these? Sure, sometimes it’s good to confirm “common sense” views, but this seems to push the envelope too far.
It’s always 8am or 8pm somewhere in the world–the graphs’ sweet spots–which is good for my e-commerce site because it’s now receiving orders from 98 countries. The shopping pattern has few peaks or valleys throughout the day, and the benefit is that orders are placed at a very regular and manageable rate. I suspect that other multinational online shopping sites have similar experiences.
For those with predictable peak shopping periods, I would recommend extra staffing in customer service at key times to answer product questions. This would include online chat (which I really like) along with phone contact. Also, server use would be predictably high during these times, so ample capacity and bandwidth would be needed along with I.T. support.
The information about online retail shopping is astounding. It’s fascinating–absolutely fascinating. In bricks and mortar, to find out the information about shopping patterns that you can about online shopping would be an enormous task, if even possible. The trick it would seem to match up the traffic/shopping data with purchase data, for targeting. I’m sure that’s considered here but maybe not.
The statistics available seem boundless. They are at any online retailer or online site’s fingertips. You are able to instantly know when they came, where they’re from, how long they looked, determine loyalty, what they looked at, how they got there and so much more. Even more astonishing, even what brought them there by device type.
Do you know everything your brick and mortar shopper looked at? Do you remember the recent discussions about the controversy of retailers asking for zip code? On line you know exactly where they came from, what they looked at by location, time they looked at it and then if they looked before. None of that is available on brick and mortar shoppers for every single traffic unit that walked through the door.
Interest leads to a purchase. Matching the patterns of view to purchase and time of purchase is key. Who’s buying what, from where, and what time and when they did–what else did they look at but didn’t purchase so that I can target them next time or review the marketing and merchandising there.
This is available. Using it–that’s the question. We have discussion all the time here about so called ‘loyalty’ cards and the wealth of data that is piled up and never mined. In online, it’s not only mined for you–it’s presented back to you in reports where you can take immediate action. Aside from matching all the purchase data to it, all the other information is free.
It’s just plain fascinating to know instantly what works, what hasn’t worked, and who’s looking at what. Fascinating. A trend of traffic by hour is barely a tip of the iceberg at the knowledge accessible from online shopping. But lets be careful to separate shopping from purchases.
It appears that limited offers, made available at odd hours seems to drive traffic–how about an “eBluelight Special?”
I’ll take the easy part of this discussion:
The data shows the very essence of the new consumer. On her terms. The easy part is describing the shift in thinking required. Every piece of retail messaging and particular promotion needs to ponder the context of the shopper and their interests/needs. In the old days a retailer needed to account for two contexts; a) how to motivate someone in the media to come to the store, and b) how to motivate a shopper to purchase when in the store. Oh my…how simple it was.
The tough part of course is to apply this to the particular circumstances of a particular brand. That’s a much larger discussion that I’ll punt on just now…but I will comment that it appears clear to me that the majority of today’s retailers still have not grasped this fundamental shift. So step one would be to test every idea against this challenge…is it based on historical retail wisdom, or is it aligned with the truly different nature of the new consumer? That alone should unearth a lot of wasted efforts and lead to some successful ones.