Guide To Device Bid Adjustments – For Services / Lead Generation Campaigns

Like most things in Google Ads, the more you look at something the more you realise how complicated it is!

The data you have available will depend on if your campaign is a lead generation campaign that gets enquiries via phone and / or a form, or an e-commerce campaign. I’ll be talking about lead gen campaigns here.

There is a pretty simple formula for working out the device bid adjustment but often there will be a lot of considerations that you need to be aware of when making your adjustments. There is a good chance that at least some of the considerations below will be effecting your account. These considerations mean that the formula on its own may not be accurate.

The standard formula for lead gen campaigns where there is no revenue data only conversion rate data is to go through each device type and set the bid to:

((Average CVR / This Device CVR) – 1) * 100

Here are the considerations – important things to consider when setting your device bid on a Lead Gen campaign:

  • What conversion data do you have. If you are tracking form fills but a large amount of conversions are via phone then you are only recording some of the data
    • usually people are more likely to call from mobile as they can just click the number. They are more likely to fill in the form on desktop as forms are clumsy on mobile. So the data you have is not representative if you aren’t recording both calls and form fills.
    • Some campaigns that aren’t well set up only record calls from Google’s Call Ad Extension. These mostly occur from mobile and aren’t very reliable, there tends to be a lot of accidental clicks on these and people accidentally calling for a competitor.
    • Some sites only track calls from mobiles, by using a click-to-call event or virtual page view in google analytics, this usually wont track any of the calls from desktop. Apart from maybe the odd caller who uses skype and clicks on the number.
  • The quality of enquiries tends to be different for different devices and for different types of enquiry (form fill vs call) so aiming for the same cost per enquiry across all enquiries and devices is unlikely to be the most profitable strategy
  • For lead gen specifically, phone calls tend to be better quality enquiries than form fills, they are more keen and more likely to convert in to a customer. Call tracking software will allow you to track desktop calls reliably.
  • Enquiries from desktop tend to be better quality than other sources. People that sit down at a computer or with a laptop are more serious than people searching on mobile. They will be doing more thorough research. So these enquiries are more valuable as they are more likely to lead to paying customers. They are often more likely to convert and are more likely to be higher paying customers. If the average conversion is worth a lot of money then desktop enquiries tend to be much better quality. Unfortunately you wont be able to see this just looking at the stats.
    • There are exceptions of course. Especially for emergency type services like emergency plumbers, locksmiths, car locksmiths, breakdown recovery.
    • For B2B campaigns it is unusual for potential customers to use mobile. Mobile users are more likely to be consumers or small businesses with the owner doing the search in his spare time. So desktop conversions are even better still for B2B.
  • Bidding – The bid you set will not be the actual bid you pay, it is the max bid, usually you will pay a bit less.
  • Statistical Significance – The data you have may not be statistically significant, if you have less than 1,000 clicks for any of the devices then the conversion rate you see will not be accurate
  • If all the campaigns are similar then it is usually safe to combine the device data at account level to get more statistically significant data. However if the campaigns are different in the following ways then it may not make sense to do so:
    • separate campaigns for different devices
    • a “call only” campaign that just has call only ads showing on mobile
    • campaigns that run at different times of day, as people usually don’t call outside of office hours
    • exclude any display / remarketing campaigns from this as they tend to get most of their clicks from mobile, sometimes this includes a lot of accidental clicks from mobile apps and websites. So mobile will have a lower CVR relative to desktop than you would find on Search Only campaigns.
    • A campaign that just has brand name terms may have different device behaviour as the brand name is used by existing customers that may just want the number so are very likely to call regardless of device.
  • Historical Changes – conversion rate by device is the most consistent part of your campaign. Whereas cost per click and cost per acquisition will have changed over time as bidding has changed, keywords change and other improvements are made. So using CVR to work out bids is the most reliable way. However any of the following changes (within the time period you are looking at) will mess up your conversion rate data:
    • changes to one of the versions of the site: desktop, tablet, phone like different design, copy, offer / call-to-action
    • landing page split tests that have taken place may have had different effects on the different types of devices. For example more wordy copy will effect what appears above and below the fold differently on different devices.
    • changes to the conversion tracking mainly:
      • which types of conversions are you tracking
      • changes to the conversion settings such as counting duplicate conversions
      • changes in analytics that may have effected conversion reporting like excluding a particular referrer. This is more likely to happen on ecommerce campaigns if the visitor leaves to go to a shopping cart and then comes back to the site. But it can happen on Lead Gen sites if someone finds a bug in the tracking that they then fix
  • Changes in bids will have different effects on volume for different devices. Especially if you are just above or below the organic results on one of the devices.
    • For some campaigns conversion rate varies depending on your ad position, so changing the bid will change the conversion rates that you used to set those bids. Usually this effect is quite small in most industries but not always.
      • If your offer on your site is a lot stronger than your competitor’s offer then you may convert more if you are the second site people visit, as they have a basis for comparison, they can see you are better. So you convert better when you are 2nd / 3rd. Whereas if you offer is weak you may convert better in 1st pulling in enquiries from people that aren’t going to bother shopping around.
      • Some industries where all competitors offer a very similar service / product the searchers that end up clicking the ads near the bottom are unlikely to convert if they haven’t already, as you are offering the exact same as the other ads, so conversion rate goes down as you lower your bid and go further down the rankings.
      • For low cost services or services people are in a rush to buy you would expect the top ads to convert better.
    • The relationship between ad position and click-through-rate is not linear. Costs go up a lot faster than click-through-rate as you get closer to the top of the page. You get diminishing returns. So if one of the devices is already in a high ad position it may not be good value to increase the bid. It would even out the cost per enquiry on the different devices but it may not work out more profitable. Vice versa an ad in a low position may get a lot more clicks for a small increase in bids.

Hopefully by considering this list you will be able to come up with a bid adjustment that is more accurate than just using the formula alone. If you are using an automated bidding strategy then you need to be especially careful as it may not factor a lot of the above.

If you want help with your campaign from a PPC Agency Wirral then feel free to give me a call.