by Alexander Dahl
Until now we have all been looking at metrics like bounce rate, click through rate and conversion rate to judge if a campaign is successful, or not. But what happens when the ad is so advanced that the potential customer can make a decision without clicking on it. Is it time for new metrics or maybe we need to evaluate the “standards” for what is a good result?
With the development of better tools and more knowledge among the designers the ads (here display ads) are becoming more advanced every day. These ads can be small flash games, ads with multiple mouse over functions or ads that link to 20, 30 or more places within a site. And this is only the beginning. Many of these ads are so advanced and contain so much information that the potential customer never need to click on the ad or enter to the website. How can the metrics we use today cope with this? And how do we analyse the results of these campaigns?
Let’s look at some examples where the results might deviate from the normal:
Example one is an ad for 10 high end apartments in an exclusive suburb placed in a national newspaper read by people of all income levels.
Here you would most likely see a large amount of impressions, but a very low click rate. If the landing page includes all the details about the apartments with text, floor plan and pictures you will most likely see a high bounce rate. People are by nature curious, so as a result many of the people clicking would need to save a lifetime in order to be able to afford the apartments in the ad. Naturally this affects the bounce rateIf the landing page is split into different sections for pictures, info, floor plan… the bounce rate might go down a bit, but it does not matter for the customer. They need people who are willing to buy. So the key metric here would be an action; to request more information. And here is the challenge: How do you judge if the amount of people requesting information is good or could be improved? If the ad as 2500000 impressions, a click through rate of less than 1%, a bounce rate of 90%, 50 people requesting information and all apartments got sold, could you say that the result was good?
In example two you have a display ad for a simple domestic product, like a screwdriver, a bread knife, or even a pair of Levi’s 501. These are products that do not need a lot of explaining. People know what they are and the most important information will be price.
As a result the difference in the performance of an ad showing the price or not will be huge.
If the ad includes the price of the product this gives the potential customer the possibility to make a decision without clicking on the ad. You get a lower amount of clicks, but your bounce rate will be lower and conversion rate will most likely be high. If the ad doesn’t show the price many people will click in order to see what it is. The result here is the opposite of the ad with the price. More clicks, higher bounce rate (as people see the price and decide) and lower conversion rate. Depending on how you look at the metrics one or the other could be better. For the client however, I think it is better to generate more traffic to the site, as this increases awareness and branding.
The last example for today is this: A client is running an online campaign with display ads. Let’s in this case use the public post service as an example. In our example they have opened a webshop and are making a campaign for their mail products: Stamps, envelopes, packaging… They have made a series of ads showing one or more of the products in each ad and the ads provide the information the customer needs to make a buying decision. The result; Lots of impressions, few clicks, but low bounce rate and a high conversion rate. The client is not too happy, because they expected a higher amount of clicks on the ads they are paying a lot of money for. However, in the time during and shortly after the campaign the amount of direct traffic to the site increases and this traffic creates further conversions. Any good marketer or analyst knows that the likeliness of a link between the sudden increase of direct traffic to the site and the campaign running is huge and as this client has a well known name and a url that most people know or can guess an ad campaign can create traffic even if the customer never clicks on the ad. How do you measure that using the existing metrics? The truth is; you can’t.
So what do you do when the standard metrics and intervals does not help you?
It requires more than just the correct ad setup and being able to put together a report with nice graphics in order to deliver an exceptional report to the client. The keys to success are understanding what the client expects from the campaign, understanding the product, understanding the potential buyer, understanding the media and the ads used. And you need to apply this insight in your reporting, explaining the campaign has performed in relation to these aspects. It is a challenge to be an analyst.
I do not think the good analysts in general are appreciated enough, because they really do make a difference for the companies where they work. So finishing off for this time let me say this: To all of you out there, making an effort and moving the world of analytics forward: Thank you! You are doing a great job!
