After a substantial period of testing, learning and cold hard results we now are in a position to fully judge Google’s newish campaign type Performance Max (aka PMax). The feedback is predominantly negative and the main theme is lack of information – it’s a black box. But it’s a really pretty, shiny black box (on the outside).
Here are the elements that we don’t like about PMax
1 – No network information: We know PMax places adds into Google Ads, Shopping (if eCommerce client), Gmail, Display Network, Discovery etc. Google refuses to present ANY data here on what spend, clicks, leads, revenue or otherwise came from which network. Of course it is their right as it is their product but if you have nothing to hide then why not provide this information? As digital marketers, info is powerful and this new ‘black box’ approach of theirs is unusual for Google and unhelpful. Hopefully this does not set a trend on how they intend to present data to marketers going toward. HACK – to find the number of impressions and clicks from the Google Shopping part of your PMax campaign, go to Google Merchant Centre and check your click volume there as it only shows Shopping data.
2 – Poor placements: Some time ago, Google very quietly added a PMax placements report found in reports / pre-defined reports / other / Performance Max campaigns placement. Here you will find your PMax ad impressions by placement website. We have seen a huge amount of wasted spend on placements like Tinder, foreign language sites, Sudoko, Solitaire and so many other gaming / irrelevant sites etc which sell display inventory to Google. Our experience shows that these sites incur copious volume of accidental clicks and thus cost to your campaign. Furthermore, this report refuses to show any conversion data…. Again as a digital marketer I am happy to spend money in these placements if they are converting. But logic and experience says they are not so they need to be cut! HACK – these placements can be excluded from the campaign set up and omitted in the settings.
3 – Search term insights: This is the equivalent of a search term report in a regular search campaign. The difference however is the infuriating presentation of this data…. The data range option is only a compare date range and you can’t choose a period. Then the table only presents conversions but not spend, clicks etc. Anyone half mathematical will know that a conversion for most clients is a waste of money if it cost you a $5000 of clicks for that conversion…. So why does Google make this so ambiguous? The lack of information is disappointing and frankly a bit deceptive.
4 – Brand spend: Segueing from the search term insights point above, we note in this table a massive amount of branded spend. If your potential customers Google your name they should see you #1 in organic results and also in the maps provided your listings there are even half way accurate. Therefore why would you PAY for this click? The reason we feel is that it makes the conversion number of this campaign look fantastic (if you don’t know better about this brand spend). HACK – block brand spend, refer below.
5 – Brand spend and SEO: With PMax claiming the branded clicks, you will definitely notice a dip in your organic traffic. So if you are running SEO, your SEO traffic will drop like a hot potato because all the brand clicks are now attributed to Google Ads.
6 – No negative keywords: Unlike search where you can block spend on wastage / ambiguous terms, negative keywords are not available in PMax. HACK – they can be added through tools / shared library / negative keyword lists which is a very clunky way of doing this – the sceptic in me thinks Google makes it hard to find and action as they don’t want you adding ‘negs’. And furthermore, at this stage you can only add negatives in this library at an account level. So if you are a car dealer that has one PMax campaign for new cars and one PMax campaign for used cars, don’t add ‘used’ as a negative keyword to help your new car campaign as it will block all used car traffic from all campaigns including used cars.
7 – No phone call information: In search campaigns, you can see a table of your phone calls from the search campaign including date, time, duration, campaign, ad group, keyword etc which is super helpful. This is found at reports / predefined reports / assets (legacy) / call details. Calls from PMax do not show in this report. This is suspicious in its absence… We have noted multiple instances of offshore PMax campaign activity and thus PMax calls may not be presented in this report as they are not genuine, local calls. Furthermore when you plug into a 3rd party phone software like WildJar, the calls are lower quality upon listening and lower duration than calls from a normal search campaign.
Here are the elements we like about PMax
1 – Results: For a select number of advertisers (particularly eCommerce clients) we are getting great results including great return on ad spend (ROAS).
2 – The network and reach: The ability of this product to find and then chase audience personas across all of Google’s properties is unquestionably awesome.
3 – Audience usage: The ability to upload 1st, 2nd and 3rd party data into the account to help Google understand your target client is super powerful. Audiences like your remarketing list, in market analytics audiences, CRM list etc help Google remarket to those people and also create similar audiences (aka Lookalike) to market to and thus expand your reach
In summary this product can get you a low cost per lead only if you understand the configuration negatives mentioned above and accept that it is a black box with poor genuine insights. From a more macro level, it also intonates that Google may be building new products in a more secretive way which will frustrate genuine data driven marketers like us! If you need help with your Performance Max campaign or search in general let us know.