Top 5 Strategies For a Successful Direct Response Campaign

Typically, when you are running a direct response (DR) online advertising campaign, you have a very limited time to prove the effectiveness of your advertising strategy and the capability of your network. If the client does not see a clear progression of performance in the right direction, they would terminate the campaign. Clients usually have 24-48 hour out clauses in their insertion order (IO). When running an online display campaign, there are lots of variables at play.  Choosing the right ones could be the difference between a successful campaign and a prematurely terminated campaign with an unhappy client.

Here are the key ingredients to a successful DR campaign:

1.       Content:  Content is very crucial component of any successful campaign. The content decides how long the user is going to be on a site, what would be the location of the ad on the page and how much will the user engage with the ad.

Consider a user on a photo sharing site. Each page on the site will have some ads. The user will most likely be skimming through pages looking at pictures spending barely seconds before moving on to the next page. If such is the case, you can imagine that the user will not even look at the ads. Contrast this with an article that is of editorial nature, a user is more likely to be on this type of page for minutes. An ad on such a page will be in front of the user for a longer time and is more likely to register and invoke a response from him or her.

The more relevant the content to the advertising campaign, the more likely it is that the user will engage with the ad. If you are selling Verizon FiOS service, for example, you are more likely to find the user on cnet.com rather than babycentral.com. Which bring me to the next item on my list namely the audience.  

2.       Audience:  When planning your campaign, it is imperative that you chose to run it in front of the right audience. As simple and as obvious as it may sound, it is often an overlooked feature of a campaign. Very often I come across campaigns that failed because they were running on mismatched audience. You can get sophisticated by employing a third party audience measuring company like Quantcast or Comscore, or you can use common sense while picking which sites to run the campaign on. Like I said above, Verizon FiOS will perform better on CNet than on babycentral. If you can get the right ad in front of the right audience you would have done the trick, unless of course you missed out on recency and frequency aspect of targeting the user.  

3.       Recency and Frequency: While deciding who to show the ad and who not to, a key decision parameter that I always use is how recently the ad was shown to the user in question and how many times has the user already seen the ad. After having run a lot of campaigns, I can tell you that you have a better chance of succeeding if you show an ad to a user who has recently seen the same ad. There is also a limit on the number of times you should show an ad to a user before deciding that the user will not engage in the future. From my experience, this number is between 3 to 5 times per day.  If the user does not engage with your ad after you have shown it to them 3 to 5 times, your chances of succeeding with this user drop significantly as you show more ads to this same user. So once you have shown the ad to as many users as you can what else can you do? You can find other relevant audience using re-targeting.

4.       Re-targeting:  Re-targeting allows you to target users who have been on the advertiser’s site either browsing or looking for more information. If you can identify such users within your network, you can be assured that these users will engage much better than other users. Whenever possible, I utilize re-targeting as a key component while building my campaigns. Performance of this audience is always superior to the others. The only issue is to find enough such users to run your entire campaign in front of them. Once you have applied all the above mentioned four techniques, you can utilize the final technique of standard targeting to fine tune your campaign and make it more efficient. 

5.       Standard Targeting: Sometimes basic targeting can save media dollars for your advertisers which could result in an efficient successful campaign. If you are showing an ad which is relevant to a certain state, there is absolutely no advantage in showing that ad to users who are outside that region. Whenever possible identify if the product is a nationally advertiser or more regional in nature. If it is regional then you should definitely geo-target the campaign to that region only. 

I have often notices that the performance of most of my campaigns usually drop off between 12 AM to 4 AM. These are the times when users are less engages, tired, trying to wrap up their work and go to bed. It is very unlikely that you will get them to engage. If you are selling a business product like FedEx Kinko’s office services, you are better off running that campaign from 7 AM to 7 PM. Also, remember you can target off of user’s time zone rather than yours. I have seen this error being made where campaign managers forget to target to user’s time zone and end up either running the campaign longer during the day or ending it prematurely for users on either side of the coast. This is an important distinction that is often overlooked.

These were, in my opinion, the most crucial components that can lead you to running a successful DR campaign. Again, not all campaigns respond positively with the same degree to all of these components. Some factors are more determinant for one campaign then others. You will have to play around with these to find out which one or ones make the most impact on your campaign. Once you identify which factors make your campaigns work, pour more traffic in those factors and reduce the traffic from others. I also recommend monitoring the campaign’s performance on a weekly basis. Sometimes factors that work initially for a campaign may stop performing as well and other non-performing factors might become dominant performers. If and when this happens you should be able to monitor it and adapt to the change.