Evaluating Your Keywords With The CPL Matrix

By , Account Manager at Hanapin Marketing

143 SHARES

Here at the PPC Hero Headquarters, we do a lot of talking about metrics. How to measure, what to measure, how often to measure, and then of course what in the world to do with all that data.

Today’s question revolves around how to make bid changes based on your desired CPL. This alludes to the Conversion Threshold Matrix, originated by one Jeffrey Allen.

As with most PPC-related questions, the answer varies. “It depends of the account” is one of my favorite answers to questions about how to manage PPC advertising, as it is terribly frustrating and yet entirely 100% true.

For a lead generation account, with a very clear idea of the CPL they are willing to pay, there is one simple table you can create to verify if a keyword should be paused, reduced in bid, left alone, or even bid up.

This table is based on the idea that you prefer to get all of your leads at the ideal CPL. But it’s understandable that some may come in a little bit above that goal. And yes, that a few conversions might come in a bit further beyond that desired CPL. And sometimes, you can make the excuse for a conversion that exceeds what you’re really able to allot for a single conversion. So how do we quantify that? With an Excel matrix of course!

The matrix below reflects those windows of acceptable variance from our desired CPL:  Screen shot 2013-05-06 at 11.27.27 AM

As you can see, my “low” CPL goal is $15. That is the actual CPL goal for this account. This matrix is built to assess those leads that are coming in past the CPL goal.  (For hints on general tactics for bid changing, check out Amanda’s tips!)

Based on the sagacious words of our own PPC Hero Jeff Allen, this matrix makes that assumption that if you must spend over your CPL goal, you should set limits to assign what behavior necessitates a bid change or bigger yet, a pause.

Let’s take a look at those keywords coming in close to your conversion goal- the first level of our matrix:Screen shot 2013-05-06 at 11.24.24 AM

This makes the assumption that if you’re coming in at 1.5x the goal CPL, you can afford up to 4 conversions. Once you’re beyond 4 conversions at that higher cost, you are going to need to take some serious measures. This will most likely look like a reduced bid.

The second row of our matrix shows that the keywords coming in above $22.50 but below $37.50 are allowed a few conversions before needing to take action. Screen shot 2013-05-06 at 11.28.34 AMThis row suggests that if a keyword is spending up to 2.5x beyond the desired CPL , more than 3 conversions is just too much.

This process continues through the third row of the matrix, and into the fourth. The fourth row of this exampleScreen shot 2013-05-06 at 11.32.02 AM

shows that if you’re spending between 3.5x and 4.5x the goal CPL, which in this case is $67.50, you can only afford a single conversion at a time. This is a big red flag for the offending keyword. Is this keyword usually responsible for some quality lead volume? If so, don’t be so hasty to pause it, perhaps you need to evaluate the other metrics contributing to this costly keyword.  What is the CTR? If it’s not looking great, reexamine that keyword’s max CPC and quality score. It may not be as simple as bidding up or down, you may need to work a little harder to make a keyword like this perform better. How is this word’s conversion rate? Maybe you’ve got a great CTR but all that activity comes to a screeching halt when the user gets to your landing page. Our PPC Heroes have tons of suggestions for how to combat issues just like this. And these scenarios are only talking about keywords you really believe in.

So let’s go through a quick application of this whole idea:

First let’s pull up a keyword report from the past 30 days. Sorting out any non-converting keywords, we’re left with those that are bringing in leads.

We then sort by           Screen shot 2013-05-06 at 1.22.37 PM

Depending on the list size and the conversion volume of your keywords, it may be most helpful to open only 1 range at a time. For this set, I’ll pull up all keywords whose CPL fell into my second row (those converting at 1.5x to 2.5x  the desired CPL). Screen shot 2013-05-06 at 12.58.55 PM

This then shows me only the keywords that have converted for the second grouping or “3 or Less.” Three is the magic number as anything more than this is just spending too much.

My results look like this:Screen shot 2013-05-06 at 1.31.40 PM

So you can see that the bottom keyword 27, 29 and 30 are going crazy! Lots and lots of conversions (which is great!) but at a cost that’s just too much to be happening so frequently when my goal is $15.

From this point, we take a look at what’s happening with these keywords. What’s the current bid? How’s the average CPC?  Does the average position give any flexibility to bid changes?

While this post is not meant to give specific solutions for every case of overspending keyword, it’s intention is to give you a measurable, repeatable threshold for making decisions about your keywords. When you’re able to plug in ranges based on your own goals, it makes that intuition of “I think I need to pull back on these keywords” something that follows a consistent pattern.  It’s a hard and fast set of parameters so that when clients ask about this whole “bid change hullabaloo,”  you’ve got more than a hunch to justify it.

The conversion threshold matrix is a structure that works for this PPC Hero, have you tried something similar? Have you found that a template like this keeps your changes consistent (not to mention easy to track)? Let us know!

ppc-hero-adwords-disapprovals_search-ads

Twitter Facebook LinkedIn Google+ Email Print More
  • http://ppccampaigngenerator.com/ Jarad Collier

    Solid advice and approach to bid management. Set thresholds of reasonable CPA (tiers) and manage based on those tiers. I like the 2x 3x CPA goal scale too.

    One thing I’ve done on a few accounts is set up automated bid rule tiers that do something very similar to what you proposed to do in Excel. For example, say a CPA goal is $15, tier 1 rule would be Cost >= $15, Conv. = 0… decrease bid by 5 %. Then set up tier 2 where Cost >= $25 and Conv. = 0, Reduce bids by 5%.

    You make as many tiers as you want.

    Then you can set them up to run daily using data from the past {whatever date range}.

    Point is, use the same date range. Let’s say a keyword has cost of $33 and conversions = 0. Since the rule runs daily, this keyword has cost greater than $15 so tier 1 bid rule will run and decrease its bid by 5%. Its cost is also greater than $25 so tier 2 bid rule will also reduce the bid by 5%. So it’s like a compounding effect.

    Anyways, just thought I’d share the idea.