Account structure is a critical element for ensuring paid search success. It ensures that keywords, ad message, and landing pages are prioritized in alignment with the business’s objectives.
However, it is equally important for structure to be as functional as possible for most effective reporting and optimization. One should think beyond the initial set up and create a nimble structure that will allow for efficient ongoing maintenance.
While some engine features vary, particularly across those in non-English speaking markets, most have the same common elements.
The below best practices focus on how to set up campaigns and ad groups. As you will see, an effective structure is deeply connected to your marketing strategy and business needs.
Campaign Setup Best Practices
When deciding what and how many campaigns to have, as a guide, use campaign settings and consider what searches you will want to have maximum coverage. Targeting and budget are the most critical ones.
For cleanest reporting, it is recommended to set up a new campaign for each combination of targeting and budget needs. This will allow to easily track how much is spent and what performance you derive from all areas.
In terms of coverage, it is a best practice to fully fund your own brand terms followed by supplier brand terms and anything that is likely to drive a high ROI.
Where Should Your Ads Show?
First, consider where do you want your ads to run. Evaluate what engine’s network you should run, in what geography and languages.
It is often assumed that ads would show up on search results (i.e. search network only). However, in some cases you may want to run on syndication sites or even the display networks of the given engine. Take a moment to check which network you need and select the right setting.
For geography, it is tempting to have the same footprint as your business does with other media. However, your search engine demand may differ from where your offline sales occur.
Take a moment to research where search engine demand is likeliest to come from (DMA, cities, states, regions, countries). Then, set up a couple individual campaigns for your top markets with.
If budget allows, also create one extra “catch all” campaign to cover remaining traffic in the rest of the possible places where your consumers may come from.
For example, a Boston retailer may create campaigns for the Boston metropolitan area, neighboring counties, New England, and the rest of the U.S.
While it is tempting to select English or all languages, do research where the bulk of the traffic will come from and if your site fully supports that. If you could be targeting multiple languages, do create multiple campaigns for each language version.
For example, if in our earlier example, the Boston-based retailer sold flags of countries around the world, its site supported three languages and it shipped only to the U.S. and Canada, its structure could look something like this.
- English_Boston metropolitan area
- English_neighboring counties
- English_rest of New England
- Spanish_New England
Notice that there isn’t a Spanish and French version to mirror each English campaign. That is due to search volume not justifying that.
Follow the Money
Before you finalize the structure, consider your budgeting and reporting needs. As mentioned earlier, there will be some terms you will want to fully fund.
For example, your own brand terms and, if you are a reseller, your supplier terms. As budget lives at the campaign level, anything needing to have dedicated budget deserves its own campaign.
Using our hypothetical retailer, assuming they are not a reseller, they should have 16 campaigns instead of the above mentioned 8 – 8 for branded terms and 8 for non-branded terms. Unless, due to special reporting requirements, budgets have to be managed against organization parameters.
Let’s imagine that there are cheap flags (with 3 or fewer colors or less that are much cheaper to produce), and the retailer wants to prioritize selling those, allocating more budget towards them. In that case, 32 campaigns would be needed
|Brand-English_Boston metropolitan area-3colors||Brand-English_Boston metropolitan area-Multi|
|Brand-English_neighboring counties-3colors||Brand-English_neighboring counties-Multi|
|Brand-English_rest of New England-3colors||Brand-English_rest of New England-Multi|
|Brand-Spanish_New England-3colors||Brand-Spanish_New England-Multi|
|Generic-English_Boston metropolitan area-3colors||Generic-English_Boston metropolitan area-Multi|
|Generic-English_neighboring counties-3colors||Generic-English_neighboring counties-Multi|
|Generic-English_rest of New England-3colors||Generic-English_rest of New England-Multi|
|Generic-Spanish_New England-3colors||Generic-Spanish_New England-Multi|
Notice how the naming convention is succinct, reflects all key aspects and uses special characters to separate them. This is key for reporting and filtering with the engine tools and any offline analysis you will do.
Ad Group Best Practices
Start by clustering your desired keywords by themes. These should be as narrow as reasonably possible to avoid overlap in targeted searches. With our flag store example, each keyword cluster could be for terms around flags of a specific country. These clusters will not necessarily translate into ad groups.
Before transforming your keyword clusters into ad groups, consider Match types, audience targeting and messaging.
For best efficiency and to simplify negative matching, it is recommended to create duplicate groups by match type. Known as match type mirroring, this best practice entails each group having only terms of one match type.
Exact Match groups should perform the best and not need negatives barring unusual circumstances. Phrase and Broad Match groups usually perform less well and are the focus for negative matching.
To note, to group cannibalization, you will need to have your Exact Match positive terms as negatives in Phrase and Broad groups.
Match type group mirroring also has budgetary benefits. At times of limited budget, it is much easier to pause less performing match types with them isolated in different groups.
If you have a very high-volume campaign and need superior control over costs, another approach can work: mirror match types by campaign so each campaign only has groups and terms of one match type.
If you want to target different user groups based on their behavior, you will need to further clone your groups based on audience targeting. As audience targeting can be also set at campaign level, this is recommended more for higher level campaigns.
It can be tempting to create a very detailed structure. If it is starting to look complex, it probably is. Consider starting a new account. Not everything needs to live within one engine account and multiple ones are often needed for larger advertising efforts, particularly with multiple sources of budget involved.
Consider also using filters and labels. These elements will allow you to quickly report on various parts of your account or campaign, without using an overly complex naming convention for campaigns and groups.
Finally, structure is not a static element. Do review it periodically especially if there are website updates. Site content and landing page changes are opportunities to improve your structure and better support the digital marketing strategy.
Featured Image: Paulo Bobita