Entries Tagged 'Analytics' ↓

A couple of days ago I had an interesting conversation about the differences between intent and interest as it applies to targeting ads.
Google is a huge business because they run the biggest “intent engine” in the world. Millions of people are going to Google every day and effectively state their current intent.
If you search for “XBOX Price” Google assumes that you intend to buy an XBOX.
The question is - Is intent all you need in order to target ads for users in the most efficient way?

I argue that the answer is no. Search is a great indicator for intent, but often enough it is not a good indicator for interest.
If you look at Google Trends you can see that a lot more users are searching for the term “Music” than the term “Britney Spears” although Britney is one of the more popular search terms. Same results if we compare the term “Movies” to “Batman” or “The Dark Knight” which was the most popular movie of the year.
We see the same thing in fashion searches as well as gaming.

When people want to buy something, they search for the exact phrase. This is why Google is a great platform for direct marketers who want to “catch” users just before they decide to buy something. But if you want to target users by their long term interests, search is not the best and definitely not the only place you need to put your money.
For example, take people who search for music. They get to iTunes, Rhapsody or any of the other music services and that’s it. Now for the next year, they will get all their music needs there. If you want to target ads for the some of your artists, how do you know which users tend to like each music genre?
Even if you use the user search data you may draw the wrong conclusions. Maybe I searched for Madonna just in order to buy a present to a friend. But if you will know which music I listened to for the past year, you can really conclude what types of music and artist I like and which ones I might like in the future.

This is the still untapped promise of behavioral targeting - the ability to go from intent to interest. It is especially suited for brand advertisers. Unlike direct marketers, brand advertisers want to reach a certain profile of users. The purpose of these campaigns is to build a long term relationship with the user and not a one night stand.
For example, they want to reach the users who are really engage into business and not the ones who looked for stocks right now because it is the main item in the news. They want to reach the real fashion enthusiasts and not the ones who searched for a dress because it’s prom night. They want to know your interests and not just what you want to buy or do today.

Now, to be fair, search could also give you indicators about interest. If every day for a few months I will search for keywords involving financial terms, we can assume that it is part of my interests. But the question is, how many people use search like that? Don’t most people search for something a couple of times, find the sites they need, bookmark them and than start to go there directly?

This is why in the long run, all these social networks, blogging and video sites might have a very bright future. While Google is sitting on today gold mine of intent data, they might be holding to the gold mine of tomorrow - the user interests data.

Microsoft announced today an important new tool that will help advertisers to measure an engagement metric for their online ads.
This is a great step forward for the industry when a giant like Microsoft (that owns aQuantive and Atlas) admits that the old way of measuring impressions and clicks is just not good enough.

The new "Engagement Mapping" tool will allow you to measure the different interactions consumers had with an ad as a way to understand its effect on the buying decision.
Basically it will take into account how may times an ad was shown across sites, and hopefully also the different interaction the users had with it, when deciding on whether it contributed to the final sale.

This is a great step forward, but you probably would not be surprised if I say MS didn’t went all the way with that.
What they are missing is also the ability to tie the different interactions, across sessions, the users had inside the advertiser site to the final decision process about the ad ROI.
Even going further, the goal of an advertisers is not always a one time sale. You actually want to measure the user interaction with your brand across time.
For example: for Amazon "Acquiring" a user that will buy one item is not as valuable as a user that will register to Amazon prime and buy at least two items every week on the course of a few months.
Also, some advertisers goal is not a sale on their online shop. A social network want to get users that will not just register to the site, but will also use it as much as possible, will have many friends in their list and upload as many photos as possible.

Still, this is a welcome step forward for the Analytics and advertising industries, and it will probably start to move other companies in the same way.

People tend to say that in the web business traffic is the king. I always add, that if traffic is the king, so ROI is god.
People sometime looks at advertising as a way to drive millions of users to your web site. True, brand awareness is very important, but I will argue that in today competitive market, what you actually want is not just a lot of traffic but the right traffic.

I’ll give you an example from our life as a company. In the recent week we got great press coverage. From famous tech blogs such as Techcrunch and Read/Write web to the mainstream news sites like cnn, forbes and others. We got huge amount of traffic and more than 4000 beta requests in less than two days.
We use our own platform, NuConomy Studio, to track the different activities on our site. One of my favorites reports in the system is the one that shows you not just how much traffic you got from any referrer to your site, but which referrer contributes the most value to you.
In our case, we actually saw that the users that comes from Techcrunch were almost 5 times more value to us than the users that got to our site from the traditional media sites (which is fascinating fact when you consider that most people who register to our beta are in CXX or VP position in their companies).

Of course the big question is how do you define value?
In our case it was the combination of number of registration, request for information and actual acceptance into our beta program.
But every business has a very different kind of definition for value. If you are a blogging platform the most valuable users are probably the ones that wrote the posts that generated the highest number of views, comments, rating, shared and clicks on ads.
If you are an e-commerce site, value is probably the users that bought the most items, spend the most money, reviewed the most products, etc…

If you an advertiser, would you rather get (and pay) a  million users click on your ad, or just two hundred thousands - but the right two hundred thousands?
If you are bidding for keywords on Google ads do you want to buy the ones that generated the most clicks or the ones that brought you the most valuable users?

I personally believe that the measurement of mare traffic is something that will slowly die. Advertisers will not be satisfied with the numbers of how much hits and uniques you have. They will want to know exactly what kind of audience are engaged in your site. They will want to know exactly what kind of ROI to expect when advertising in your site.