A Guide To Auction Insight Analysis: Part 3
Final tips & tricks
Now we’ve looked at accessing the data and analysing the output, this final entry will run through some tips and tricks to keep in mind when working with auction insights data.
Always keep in mind changes to the account that can impact your performance in the auction, to help avoid spurious conclusions.
This could be, for example, a tROAS change resulting in a drop in ad rank opening the gate to competitors - rather than competitors becoming more aggressive, you have become less competitive.
You can quickly check this by cross-referencing the change history tab with your analysis dates:
Reviewing the full range performance metrics alongside your auction data also helps provide context. Keep in mind the total impressions you’re seeing, your CPC etc.
Lastly, don’t forget to consider the impact of external factors such as holidays or events on competitor behaviour (obvious ones include Black Friday or Christmas, or more vertical specific events like university clearing or half term)
Don’t forget that Google provides a device segment. This can help spot where a competitor’s strategy involves focusing on a specific device.
This heavily leans into competitor research as a follow up - if you do find a competitor who is opting out of desktop and focusing on mobile, the next step is to look into why that might be.
Be aware that when exporting data including device-specific metrics, the percentages provided will be specific to each device. Combining them to calculate an average or total will not accurately reflect the overall average across all devices, resulting in an inaccurate calculation.
Auction insights only gives you data for auctions you were eligible to show - not an absolute total.
Your best bet for an accurate view are campaigns without budget caps (or keywords in these campaigns) and a competitive impression share across the day, otherwise you may just be looking at a slice of the landscape:
The below two charts highlight this pitfall - this particular campaign was hit by a delivery issue that resulted in very few impressions being served at the start of the month, despite strong impression share. As shown, once resolved we started to get an accurate view of the competitive landscape:
Be aware of the impression share threshold that determines which URL domains are included in the report.
While you may see a smaller competitor over the last seven days, they may disappear when looking at the same view over the last 30 days.
This is particularly important for analysis over longer timeframes - looking at a 6-12 month period may mean you miss out on a new competitor who has only just entered the auction but is making serious waves.
In the same vein, you may see more competitors in report when looking at keyword or ad group level vs campaign level.
Use the right level of granularity depending on data availability (the report will not generate if the element does not reach >10% impression share) or what your analysis is trying to achieve (are you looking at a deep dive into your top generic keyword, or looking for a snapshot of competition across the account)
Across the Impression Share metric, valid values range from <10% through to 100%
<10% is difficult to work with as there is no way of knowing if this domain is at 1.9% or 9.9% impression share. Keep in mind if working in Sheets or Excel, that this will impact your sorting as <10% will be recognized as a text rather than a number (example below using TYPE formula):
Auction Insights does not provide any detail on how a competitor entered the auction, just that they were present.
With an increased use of broad match across accounts, those who appear to be running competitor campaigns on your brand could be matched via generic keywords, rather than an intentional move to target your brand.
If you want some additional detail, you may be able to get this from their landing page query string if campaign names are exposed through human-readable parameters such as UTMs:
https://www.mycompetitordomain.com/landing-page?utm_source=google&utm_medium=cpc&utm_campaign=competitor
Auction insights can be used to support analysis through third party competitor research tools, such as SEMRush or Adthena.
Auction insights data is inherently more reliable as it's backend data Google is sharing with advertisers, the main drawback being the limited scope of what we can access.
Try using auction insights to inform what competitors you should be looking at (for example, those who appear in the aggressive segment in your competitive matrix discussed in part two), and take those domains into these third-party tools to get a broader (but inherently less accurate) view of their tactics.
You may see some URL abbreviations, especially if special characters are used in a URL.
By default, your domain will come through as You under the display URL domain column
Hyphens are a good example, a domain like my-website.com would come through as website.com in the display URL domain column.
Agency specific, do involve your client team with your analysis - especially those who may be involved in broader competitor research or market positioning.
This can add another layer of context, knowing what white label agreements are in place, who are in talks to buy who etc.
Lastly, use tools such Google Ad Transparency Center to bring your competitor analysis to life.
This is a great way to quickly get sight on the range of messaging a competitor is using, and it’s completely free to use:
And that’s it for this series! Hopefully it has been helpful in providing some structure for those just getting started with auction insights analysis, or a refresher for those who are jumping back into the data.