Last week we organized a happy hour here at our office in New York. This was called & # 39; Energizing Ecommerce: Retail winners, losers and Amazon & # 39; and was made in collaboration with Jumpshot. Here are highlights from the Amazon and Google event, and their market shares in e-commerce.
Our speaker for the evening was Steve Kraus, Jumpshot & # 39; s Head of Digital Insights. Steve has been with Jumpshot since November 2018 and previously held positions as Chief of Insights at SimilarWeb and SVP and Chief Insights Officer at Ipsos. He obtained his PhD at Harvard in social psychology and was previously a professor.
He added their data to consumer digital behavior to ask and answer various questions we all have about how consumers search and spend online.
Because they work primarily with clickstream data, the numbers in this article refer to clicks on the desktop and mobile devices, but in-app promotions are excluded. (You can read more about their details at the bottom of this article.)
What is Amazon's market share in US digital transactions?
Steve immediately jumped in with a question that we all probably asked at some point: how much does Amazon matter in digital transactions?
Many in the audience estimate a share of around 35-60%. But in reality the distribution looks like this:
Amazon has an average share of 74% in digital transactions in the US.
In this graph, the Y-axis represents the number of digital transactions and the X-axis represents Amazon's market share in each. So for a category like electronics, you can see that there is a huge volume of digital transactions and that Amazon might hold 80-85% of that volume. For something like women's clothing, on the other hand, both the volume and the market share of Amazon are smaller – closer to 35% (which we will discuss later).
Steve said it's pretty hard to capture every interaction in the long tail, but the numbers are still a fairly accurate representation of the total.
The most important takeaway meals?
"Everyone says Amazon is that big, but in reality many of us still underestimate it."
According to Steve, people may not hypnotize the dominance of Amazon (nor that of Google). And they can even underestimate it.
"Many people estimate Amazon's market share at around 40-50% – but that's how they perform in their worst categories, such as clothing and furniture."
He has expanded and added this point Amazon has 16x the number of transactions from Walmart and 54x the number of transactions from Target.
"Although we hear so much in the news about antitrust, congress hearings, etc., we still underestimate how big and powerful (these technical giants) really are."
What percentage of searches takes place on Google?
Another very popular question: How much of the search market does Google own?
According to Jumpshot's clickstream data, Google has around 90% of all searches on the internet (desktop and Android phones).
The big disadvantage here is that they cannot watch searches that occur through voice searches or in an app – YouTube, Google Maps, Amazon and others probably all have huge volumes of search activity that are not taken into account. For example, the chart below for YouTube seems much smaller than Google Images, and we assume that it is at least partially because most Google Image users use desktop and most YouTube users use the app.
We can also see from this chart how, overall, since January 2016, the volume of desktop searches has decreased in all areas (and we assume that the volume will be moved to mobile).
This graph shows the same data as above, but with Google removed so that we have a better view of the smaller players.
An interesting point to note is how much DuckDuckGo is climbing. However, their total volume remains relatively small. They have gone from nearly to 300 million searches per month, which is great. But compared to Google's 60 billion searches per month, they only reached 0.5% of the volume.
What percentage of product-related search takes place on Amazon?
We have seen that Google holds the largest share in the search market. But what if we look specifically at product-related search?
Jumpshot discovered that the positions of Amazon and Google were reversed between 2015 and 2018. As shown in the image below, Google had 54% of product-related search in 2015 and Amazon 46%. In 2018, Amazon had risen to 54% and Google had fallen to 46%.
The obvious concern here, and which Steve has demonstrated through the Vox header, is the advantage that Amazon has of all the data collected in those searches. And of course, how little of that data is shared with sellers on Amazon who are made to compete with Amazon's own product lines.
By simply researching "what kind of keywords people are looking for and not conversing?", Amazon can discover market opportunities and enter them in ways that their competition cannot.
The story here, "Amazon has created a skincare line based on what users are looking for," is an example of the qualities of many sellers, regulators and consumers.
Which retailers are successful and which are struggling?
In the Amazon and Google era, which retailers seem to figure out how to stay competitive?
Contrary to what is often claimed about the & # 39; retail apocalypse & # 39 ;, Jumpshot says that many traditional physical retailers are doing pretty well. Other newer digital brands are too – Chewy and Wayfair for example.
This graph shows the distribution of a number of successful and struggling retailers. The y-axis represents the total number of digital transactions of that brand in 2018, and the x-axis represents the change in their number of digital transactions between 2017 and 2018.
All names that are displayed in green therefore increase their number of digital transactions. Brands shrink in red.
According to Steve, the retail apocalypse is a & # 39; story that is over-sold & # 39 ;.
On the contrary, he says that "many brands that are known for their presence in the retail sector are finding out omnichannel."
What makes sure that certain retailers succeed and others do not? According to Steve, part of the reason can be traced to the type of search: brand versus utilitarian.
Branded versus utilitarian searches
There are two very different types of searches in the world, says Steve.
- Branded searches, for the name of a certain brand such as Nike, Bose or Nintendo
- Utility or unbranded searches for items such as & # 39; phone case & # 39 ;, & # 39; men's socks & # 39; and & # 39; paper towels & # 39;
And as we can expect, Google and Amazon play very different roles in these two types of searches. In the graph below we see search volumes for the top 1000 brand and non-brand keywords. Google's volume is on the y-axis and Amazon & # 39; s on the x-axis. Items are in blue brand searches, items in orange are unbranded searches.
Interestingly, both Prime Video and Audible get more search volume on Google than on Amazon (despite the fact that they are owned by Amazon).
Terms such as Bluetooth headphones, hard drives, paper towels and phone cases receive very good volumes on both Amazon and Google.
It is remarkable here that it is difficult to distinguish whether some of these conditions are product-related or knowledge-related. If someone searches for & # 39; iPhone & # 39 ;, does he or she want to buy an iPhone, learn about the next iPhone to be released or get help from the person he already has?
Steve also pointed out that furniture is another interesting category. Wayfair has had so much success there to increase the market share of the search function. But a few years ago we might have predicted that it would be a category that Amazon would have dominated: large ticket items where peoplehave no strong brand preferences. Surprisingly, Amazon has not done so well there.
If we look specifically at clothing searches, we see that they too tend to fall under brand name or utilitarian.
Amazon usually does well on unbranded products such as socks and underwear. For example, Heinz can be very successful on Amazon. Larger, more famous brands, on the other hand, tend to see more traffic on Google.
In the image below, Jumpshot noted that "large fashion brands would be at the top right." For example, they say that Nike has 11 million searches on Google versus only 250,000 on Amazon.
To be competitive in this context, Steve might ask:
“Which search terms do not grow and convert? That is the real sweet spot of opportunity. "
Organic clicks, paid clicks and zero clicks
There are three possible actions after a search: organic clicks, paid clicks or zero clicks.
Another question many of us ask is, "Which of these, if any, is on the rise?"
The graph below shows trends of click types from January 2016 to March 2019. Paid clicks and zero clicks rise and organic clicks fall.
And of course, the whole idea of zero-click searches has many people asking questions.
- What percentage of search queries on Google don't click?
- How many people now find what they need on the homepage?
- What does it mean for the publishing industry if the content that a company creates can deliver information to a user without that user ever leaving the Google site?
- How much busier and more informative will SERP & # 39; s become?
Just like Amazon, Google has extremely valuable search data. It is demonstrable in their own business interest not to share it.
How do people find things online?
Discover -> search -> shopping -> buy.
In this image, Jumpshot shows the funnel of the consumer and how Google and Amazon play dominant roles in the search and shopping phase. From the & # 39; search & # 39; Google has almost 90% of the total search volume and Amazon has around 54% of the product search volume. Amazon has a share of around 75% of all categories in the "shop" category.
However, it is interesting that the upper part of the funnel is a little more obvious. And as Steve points out, it's a very difficult thing to quantify. We can quantify if you search or shop, but discovery has less immediate action. "
We can assume that most discoveries take place on sites such as Facebook, Instagram and Pinterest. And while Facebook was perhaps very strong there, they made changes to put less emphasis on advertisers and on their way to privacy.
At one point a member of the public asked Steve about his thoughts on regulation and privacy.
"My own suspicion is that here in the US it will not happen soon. In the EU yes, it is already happening. But in the US it seems almost & # 39; un-American & # 39; to talk about antitrust. These are entrepreneurs. They made their own success. Why should we punish them? I don't see the political will to do that here. "
At the same time, he noted that "you see that Facebook and Google are reluctantly giving up some things to stay out of the way."
What kind of data does Jumpshot look at?
Jumpshot works with anonymized, aggregated, clickstream data from around 100 million devices, including desktop, mobile internet for Android and app use. (Keep in mind that their app data is in a separate data set and generally does not show what happens in an app. It shows earlier whether an app has been opened, how often it has been opened and whether a user has clicked through to something else.)
With clickstream data, Jumpshot can only look at which websites people have visited. They can model age and gender based on those websites, but in general their display is quite high.
She also calibrate that data against "known sources of truth" such as actual versus predicted visits to a particular website. Based on those known sources of truth, they can develop algorithms to calibrate their numbers to those actual known numbers.
In general, Steve said, their data is most accurate when they have more known sources of truth (such as the American market).
"When Google and Amazon are so dominant, what does that mean for brands?"
We asked Steve what he found surprising or interesting about these findings, and he said, "In this world where Google and Amazon are so dominant, what does that mean for brands? How do they thrive? How do they make decisions of utilitarian versus brand economy? They have to decide which one to play in. Some brands are strong enough to do both. "
He worked out the example of Nike. They came out a few years ago with a controversial announcement that they were going to sell directly on Amazon, which they had not done before.
Even if a brand does not sell directly on Amazon, the likely scenario is that they are still there. Consumers can still go to Amazon, search for Nike and receive results from resellers and wholesalers. Products may or may not be counterfeit and sellers may or may not offer good service.
"A brand that is not on Amazon actually relinquishes its presence on Amazon to fewer people, to those market forces they have no control over," says Steve.
Nike wanted to control that, so they started a complex process to find out what and how they were going to sell on Amazon. They clearly wanted to keep their nike.com site and belong to the brands that are strong enough to do that. So they started to find out what consumers are looking for on their site versus on Amazon, and which products and prices have to offer on each of them.
For retailers, Steve says, the question is:
“Do you want to work with Amazon, via Amazon, around Amazon? How do you thrive in this digital world where so many aspects of it are these & # 39; winner-take-all & # 39; markets? & # 39;
Five extensions that directly complement the SEO strategy on the page, help you achieve rankings, track the performance of your website and measure the most important online statistics.
It is important to focus on one or two tests to ensure that you achieve statistical significance. Manual for testing ad texts, landing pages & new search engines and ad formats.
During its long history as a crucial factor in the ranking, PageRank was surrounded by a multitude of misconceptions. Five of the most common have been addressed.
There are five key elements behind a strong sales funnel. Focus on awareness of your company. More tips included optimizing sales judges.