A successful campaign can be measured by monitoring three key performance indicators from your RetargetLinks dashboards.
Check the number of clicks on the links
The first indication of a successful campaign can be determined by looking at the number of link-clicks on your Links Dashboard.
If you’re sharing the right content, to the right audience on the right channels, you’ll have a lot of clicks. The example below is from the Traction Conference Link Dashboard. They managed to get 85,138 clicks (58,296 unique) from 873 links shared via their email newsletter (Direct), Twitter and Facebook.
Check number of ads being displayed
You can measure the second performance indicator by looking at the number of ad clicks on your Ads Dashboard.
If you’re displaying relevant, compelling banner ads (good messaging, clear CTA, some animation) you’ll catch the attention of your targets and they will click on them. Traction Conference managed to display 161,340 impressions to most of the 58,138 people that clicked on their short links. Out of those, 422 people clicked for a 0.26% click-through rate (x3 times the 0.10% average for banner ad performance).
Monitor landings and conversions
The final indication of successful campaign performance can be determined by looking at the number of people that land on your page and the number of those that convert.
In the case of Traction Conference, 947 people landed in the marketing page (422 that clicked the ads and 525 that came to the site via other means, usually via a Google search) and 186 that bought a ticket for the conference. The conversion ratio is a very good 20% rate (x10 times more than regular conversion rates).
What happens if I am over-budgeting?
You can never be over-budget. This is because the system shows a maximum of 15 ads per person that has been “cookied”. We achieve this by implementing frequency caps (2 an hour, 5 a day, 9 a week); based on average browsing activity, it takes about 2 weeks on average to display 15 ads to any person.
So if you have tagged 1,000 people, then you will display 15,000 ads (15 / person) over 2 weeks. Your weekly impressions will be 15,000 ads / 2 weeks = 7,500 ads per week.
Assuming you are paying $8 CPM (per 1,000 impressions), then the weekly budget should be 7,500 ads / week x $8 / 1,000 ads = $60 / week.
If your budget is $100 / week, then your cost will be $60 at the end of the campaign ($40 will be carried over).
What happens if I am under-budgeting?
Take the same example as above. Let’s suppose that you have allocated $25 a week to show your ads to 1,000 people you have tagged.
Assume the campaign last two weeks or that you keep retargeting 1,000 people every two weeks. You are effectively allocating $25 or 3,125 ads ($25 x 1,000 ads / $8) to target 1,000 people.
This is equivalent to showing three ads per person, which is not enough for the campaign to be effective. Studies have shown that it requires an average of 10 ads for someone to decide to take action. Less than 5 ads will achieve nearly zero results and diminishing returns occur when showing more than 15 ads (hence our default maximum).
This means that you will be wasting your marketing budget if you are showing less than 5 ads per person so it is important to get the budget right.
In the example below, spikes in the number of ads shown per day indicates optimal budget. Frequency caps cause the spikes based on the natural browsing activity of the people you have tagged.
You can increase your banner ad performance by identifying the “plateaus,” by increasing the ad display budget limit to support the growth in impressions.
There are different metrics to monitor when launching a link retargeting campaign.
First, keep a close eye on tagging performance of your links in the Links Dashboard. Next, track ad impression performance in your Ads Dashboard. More advanced metrics like click-through rate (which is normally between 0.30 and 1% in our system). In addition, landings and conversions are also important.
Lastly, budget your campaign correctly by avoiding plateaus in your impressions metrics.