Determining The When And How Of Cross-Channel Budget Allocation

By Jeff Allen | @JeffAllenUT | President at Hanapin Marketing

Google’s global search market share is 67%. That volume combined with the decent quality of traffic Google provides is the reason why so many PPC advertisers spend 75% of their PPC budget with the engine. But how do you know when it’s time to branch out?


Are You Budget Constrained And Running Only (Or Mostly) In Google?


Break your campaigns and/or keywords into buckets. Bucket 1 is all your top keywords that drive very high returns. These keywords could be brand terms if you bid on those, it could be exact match for your top volume drivers, etc. These are the keywords that I would make sure to run in Bing. Not that the return is going to be the same, but you want to see if the return on that bucket is better than your return on bucket 2 in Google.


Bucket 2 consists of all the mid-tier campaigns and keywords that meet goals in Google. If you have money left in the budget, you’ll run these in Google and test them in Bing.


If you still have budget left over, there is bucket 3. These are the keywords that drive volume but at poor returns, or don’t drive volume, but you leave them running because every so often you receive an incremental conversion. If you run this campaign/keyword in Google and still have budget left over, you can run them in Bing as well. But if you are budget constrained, it is unlikely you’ll have to spend on bucket 3 keywords at all!


Are You Budget Constrained But Want To Be In Paid Search And Paid Social?


This is a bit trickier, especially if you aren’t already running campaigns in all the engines you want. That’s because you’ll have to do some testing to find out what the ideal marketing mix is.


We’ll tackle the easy one first. If you are already testing in all the engines you want to be in, you still apply the three-bucket method!


1) I’d get all your campaigns in one spreadsheet along with performance stats for a 30-day period (whatever stats you like to review).


2) Sort the campaigns by your primary KPI. That could be revenue, ROAS, CPL, cost, etc.


3) Find the point at which the sum of the campaign spends is equal to your monthly budget and pause the campaigns below that point. (Meaning, if you are running 50 campaigns, and you sum the first 30 and that equals your total monthly budget, you’re going to pause the bottom 20 campaigns.)


4) Leave the rest running!


Now, this strategy assumes two things. The first is that your top campaigns aren’t limited by budget. If they are, you should open those up! Then redo this process after you collect some more data. (7 days will give you a directional trend if you don’t want to wait the full month).


The second assumption is that your campaigns are structured in a way that your top performing keywords and targets are in campaigns together. There are many instances where this isn’t the case, so you could do this at an ad group level or do a restructure before doing this process. I prefer to have my best keywords in their own campaigns when budget constrained specifically for the purpose of proper budget allocation!


Now for the more difficult issue – when you are thinking about expanding into new networks that you haven’t yet explored. This requires some forethought, testing, and patience. It also works for both budget-constrained folks and those that can spend as much as they want as long as they hit ROAS goals.




Who’s your audience? Are they male or female? Do they shop online or in stores? Do they know about your product or are you trying to increase awareness? These questions, and about 50 more, are the types of things you have to ask yourself. Paid Search advertisers tend not to have to think about this as much because they are bidding on keywords, which tend to have a whole lot of intent. But branching into new engines, like Facebook and Pinterest, requires a deeper understanding of your audience.


Once you’ve thought this through, and hopefully used data to support, you figure out which engines make sense. How do you know what engines make sense? Compare your demographic and psychographic data to the data available on engines. I did a quick search and found one article that will probably answer 80% of your questions.




Now, you have to actually launch campaigns and see how the return is. If you can’t do all of them at once because of budget, I would prioritize in this way:


  1. Google
  2. Bing
  3. Facebook
  4. Programmatic
  5. Pinterest
  6. Twitter
  7. LinkedIn


That’s just my general take, but of course, you might not want to be in all those engines, or your demographic data may tell you one is better than the other!




Since these are new channels, it is going to take some time to figure out your return. My very general rule is that you need 2 weeks of data to make optimizations and that it takes 3-4 rounds of optimizations to get things running well. So that’s 6-8 weeks of running before you know what campaigns are going to fall into which of your three buckets. But once you do, you apply the same logic of getting all your campaign data in one spreadsheet, sorting by your main KPI, and adjusting which campaigns are active and paused based on the data.