As marketers we all rely on some form of AdWords conversion tracking to optimize campaign performance.
We get our heads stuck in data for hours trying to make the best possible decision. We argue over low-performing keywords with clients and colleagues.
In the end our decisions are based on conversion data that is reported inside of Google AdWords.
But in our day-to-day work we all forget one important factor: conversion data is always biased and incomplete.
This article highlights seven major factors that have a big impact on performance.
Google AdWords Attribution Models
Attribution Models (Source: AdWords Help)
The default attribution model for AdWords conversion tracking is last-click attribution (you can now change that). That means if there are multiple AdWords clicks before a conversion, only the last click will get conversion credit.
Based on last-click attribution keywords which are at the beginning of the research process will tend to show a higher cost per conversion. This is because people might initially find your website through a broad search term but then refine their searches further as they become more informed about what they actually want.
Very specific keywords are typically close to the buying decision and are therefore more likely to be the last click before the purchase. These keywords will show a better performance based on last-click attribution.
If you are running campaigns targeting generic keywords (“car repair” etc.) as well as a campaign targeting your own brand name (“tom’s car repair”), the brand campaign can “steal” some of the credit that should actually go to the generic campaigns.
Check out the AdWords attribution tool to understand the implications of different attribution models on your cost per conversion (in your AdWords account, click on Tools -> Attribution).
AdWords Conversion Tracking vs Google Analytics Import
There are two main ways to track ad performance in Google AdWords:
AdWords conversion tracking
Importing goals or transactions from Google Analytics
Most marketers do not realize that there are significant differences between AdWords tracking and imported Google Analytics goals/transactions. Two of the main differences relate to conversion attribution and date of attribution.
Both AdWords and Google Analytics use last-click attribution. The main difference is that Google Analytics uses last-click attribution across all channels whereas AdWords tracking does not consider other channels.
This means that if there are multiple AdWords clicks before the conversion, AdWords will credit the last click with the conversion. However, Google Analytics will look at all traffic sources (including organic traffic, referrals etc.) and credit the source that generate the last click before the conversion. So even if there were five AdWords clicks before a conversion, Google Analytics will only give AdWords credit if the last click before the conversion came from AdWords.
In practice, this means that importing Google Analytics goals/transactions into AdWords will often result in a lower number of reported conversions.
Date of Attribution
Google Analytics records conversions at the actual date and time when they occur. On the other hand, AdWords credits the conversion to the date of the last click. So if someone clicks on an AdWords ad on March 19 and and returns to the online on March 23 to eventually buy a sweater, Analytics will record the conversion for March 23 whereas AdWords would record it on March 19 (when the last ad click happened).
One of the major consequences of this is that conversion numbers based on AdWords tracking can still increase retrospectively for as long as you set the cookie duration. If you check last week’s conversion numbers today and the again in 30 days from now, you will typically see the number of conversions has increased due to the fact that AdWords credits the last click not actual transaction date with the conversion. The additional conversions that happened in the meantime are referred to as post-conversions.
For a full comparison of the two tracking systems check out these Google articles:
How Cookie Duration Affects AdWords Performance
The standard cookie duration for AdWords conversion tracking is 30 days. In other words, conversions that happen within 30 days after an AdWords click will be credited to that last click. The cookie duration can be set anywhere between 7 and 90 days. Conversions that happen after the cookie window will not be credited anymore.
Here’s an example why it’s important to choose the right cookie duration for your business: While working with one of my eCommerce clients, we realized that a significant amount of conversions happened near the end of the standard cookie window. After increasing the cookie window from 30 days to 90 days, we ended up tracking an additional 20% conversions.
To set the right cookie duration for your business, check the ‘time lag’ report in your AdWords account. From your AdWords interface, click on Tools -> Attribution -> Time Lag.
The example above shows an average time lag of 5.88 days to a conversion. In this case the standard cookie duration of 30 days would be sufficient.
If your time lag report shows a lot of conversions near the end of the standard cookie window, make sure to increase the cookie duration. The longer you choose your cookie duration the more post-conversions AdWords will tend to credit.
Check the next point for more details on post-conversions.
Implications of Post-Conversions
As discussed earlier, standard AdWords tracking will credit the last ad click with a conversion. As a result, AdWords conversion numbers can increase retrospectively. That means if you check your conversion numbers for last week today and then check the same time period in 30 days from now, the conversion numbers will normally have increased.
These additional ‘post-conversions’ obviously didn’t actually happen during the week that you initially checked but the last AdWords click prior to the conversions happened during the same week. The amount of post-conversions depends both on your cookie duration and the type of product or service you are advertising.
A longer conversion time lag and cookie duration implies a higher amount of post-conversions in your AdWords account. Therefore you need to keep in mind that AdWords performance data for any given time period is only final after cookie duration. So in most cases 30 days from the last day of the time period you are looking at.
Implications of Phone Sales
By default AdWords doesn’t track phone sales or enquiries. That means AdWords doesn’t get any credit for any conversions that happen on the phone.
Depending on the business, phone sales/enquiries can make up 20%, 30% or even up to 50% of the total conversions. You need to take this into account when you check the cost per conversion in your AdWords campaigns.
There are plenty of phone tracking solutions out there that integrate with Google AdWords. What you can do is implement one of them for a couple of months to find out the ratio of phone conversions to online conversions. The ratio is typically very stable so once you have determined it you can just give AdWords the extra credit for it.
When we analyse the effect of our AdWords optimizations we typically do a before-and-after comparison.
AdWords performance: compare the same time periods whenever possible.
The right way to do a before-and-after comparison is to compare the same time period across years, i.e. May 2015 vs May 2016. The wrong way to do it is to compare last week to this week or April 2016 vs May 2016. This is because all businesses have an underlying seasonality that affects performance throughout the year (typically more than any other factor).
Second, when you are comparing time periods you need to ask yourself whether the tracking setup was the same. Where you tracking the same type of conversion? Was the cookie duration the same? Has the tracking setup changed from AdWords to Analytics import? Just one change in the way you track performance will make a performance comparison like this unfeasible.
Actual Revenues vs Tracked Revenues
How to spot changes and irregularities in your tracking setup over time
This article covered a lot of different areas that can change your conversion data significantly, This last point is not a factor impacting performance but rather a tip to spot changes or irregularities in your tracking setup over time. A great way to assess accuracy and changes in your conversion data is to compare it to hard sales figures.
The screenshot above shows an example of how to do it:
- Create a spreadsheet and add your actual monthly revenue in one column.
- Then add two columns for revenues tracked in Google Analytics and Google AdWords. Calculate the percentage of tracked revenue to total revenues next to each of them.
Obviously these ratios will be nowhere near 100% because AdWords doesn’t generate all of your revenue and because not all of the sales happen online. However, the ratios over time can easily highlight if there was a problem or change in your tracking.
Final Tips and Thoughts
We have discussed various factors that can change your AdWords performance data significantly: attribution models, tracking setup, cookie duration, phone sales, post-conversions etc.
Here are a few specific examples:
Last click attribution will show different results than first click attribution
AdWords tracking will report different results than imported goals/transaction from Google Analytics
A longer cookie duration can increase (i.e. improve) your AdWords performance by capturing additional conversions
A sudden shift towards more offline conversions (e.g. phone sales) can make your AdWords performance look worse
AdWords conversion data can improve retrospectively due to post-conversion
So what’s the main take-away here?
- Keep things in perspective. Remind yourself that AdWords conversion data is based on your specific setup and often doesn’t show the full picture.
- Keep track of any major changes with respect to your tracking and settings.
So when you have to make an important AdWords decision, take into account external factors and look at your data from perspectives using the AdWords attribution tool.
About Stefan Maescher
Stefan Maescher is an advertising professional and blogger at stefanmaescher.com. He helps businesses maximize their ROI online by using data-driven optimization strategies. He also teaches his best insights in an online course called AdWords Profits Shortcuts.