One of the greatest gifts digital marketing represents is the date to know when we are achieving our goals.
The most straightforward measurement is conversions.
Conversions can track any action a brand will find valuable from a sale to spending a certain amount of time on a page.
Yet not all brands are capable of using conversions. There are a number of reasons for this:
- Web host doesn’t support conversion tracking code.
- Brand doesn’t have on-demand access to a technician who can install the code.
- The ad network doesn’t support conversion tracking.
In these cases, a brand may have to rely on metrics outside of trackable ROI and ROAS.
As you develop your strategy to measure success, be sure you account for any limitations and how those might impact how you scope out your KPIs.
Measuring Success When You Have Access to Conversion Data
We’ll start with the best-case scenario – your brand has the ability to track conversions and can leverage that data in automation, reporting, and scaling.
Step 1: Choosing the Right Conversion Action Settings
Ad networks will ask you the following when configuring your conversions:
- What action do you want to track?
- How long is the conversion window?
- What is the conversion action worth?
- Do you want to track every conversion a user completes, or the unique user?
- What attribution model do you want to use?
For most brands, the main action you’ll track will be landing on a specified page.
This page will be determined by the code living on it or specifying the URL for your tracking parameters.
This action requires the URL to change when the user hits the order/form submission page.
If your site uses iFrame forms or the URL doesn’t change, you will need to configure an event conversion action.
On conversion window length – there is no right or wrong answer here.
Selecting the right one for your industry and your specific brand is crucial to ensure you don’t miss valuable visitors.
Yet setting the conversion window too long can create false positives and give extra credit to an ad.
If you’re unsure what’s right for you, going with the default (30 days) is safe to start with.
Brands with longer sales cycles may want to start with a 90-day window, particularly if sales take more than two quarters to complete.
Setting conversion values may seem difficult if you’re in lead gen but it’s a crucial step if you’re going to hold your accounts accountable for ROI/ROAS.
When deciding how much a conversion is worth, consider the following:
- Is the sale complete after this conversion?
- If not, how good is the resource receiving this lead at closing the deal (your overall conversion rate)?
- What is the sale worth when completed?
For example, a lawyer may value a case at $20,000, but they also may have a lead to case rate of 30% (as in for every 10 interested folks that get in touch, only 3 will actually be a case they can take on).
They also need to consider whether calls, form-fills, or chat systems provide a higher quality lead.
They may value a phone lead at $500, because their in-take team does a great job of closing the deal, while a form-fill at $50 because of historic performance.
This is why it’s important to set up conversion actions for each major type of prospect interaction – you don’t want to lump high and low value prospects together.
Nothing hurts more than double-counting your conversions. Unless you’re in auto, travel, real estate, or ecommerce, it is unlikely that your brand will get value out of tracking every conversion a user completes.
Most will want to count “one” conversion action per user.
The final consideration is deciding on an attribution model.
By default, you will be opted into “last click”, which means only the last entity the user engaged with before converting will get credit.
There’s also “first click” which will only give credit to the first engagement in the customer journey.
While the “best” attribution model (data-driven) is a great goal to shoot for, it requires considerable conversion data: 15,000 clicks and 600 conversions in a 30-day period.
So that leaves us with “manual” fractional conversions:
- Time Decay: The further away the conversion action means the less credit the action will receive.
- Linear: Every engagement gets equal credit.
- Position based: First and last engagement get 40% of the credit each and all other steps share 20%.
It’s especially important that you match your objective with the right attribution model. If not, you might end up with super successful campaigns “under delivering” because they didn’t get credit.
Step 2: Establish Benchmark Metrics (Current State)
It can be tempting to latch onto “industry averages”, but your business will have unique market factors that must come first.
Population density, competition, and age of the account can encourage more aggressive or less aggressive benchmarks.
If the account is “new” (under 90 days old) wait till you have at least a full quarter under your belt before setting the benchmarks.
Metrics to focus on:
CTR to Conversion Rate
If CTR is high and conversion rates are low, that means your landing pages are off, or the quality of traffic might be off.
If CTR is low, and conversion rates are high, that means ad copy can be improved.
CPA vs Conversion Quality/Value
When CPAs are low and the conversion quality is low, that means you have a false positive.
If the CPA is high and conversion value is low, that means there’s a structure/conversion counting problem.
It’s important to note “high”/”low” CPAs are subjective to industries. For example, I wouldn’t bat an eye at paying between $200-$500 CPA for legal, but would be in ultra optimization mode if that was my fitness CPA.
Return on ad spend may seem like an ecommerce only metric (due to needing conversion values), but it’s actually applicable to lead gen as well!
If your campaigns can’t deliver at least a 200%-300% ROAS, odds are there’s a structural issue or you might be going after a part of the business that doesn’t warrant PPC.
Impression share (and where it’s lost) vs. auction prices of converting terms (search only)
Not all keywords are equal and if a valuable term is languishing due to rank (underbidding) or budget, you’ll want to prioritize it or reconfigure your automated bidding.
Step 3: Set Realistic Growth Goals
Once we’ve established your core metrics and benchmarks, it’s time to challenge ourselves to grow!
Goals should be established at month over month, quarter over quarter, and year over year.
The smaller the budget, the greater percentage improvements you should hold yourself accountable to.
For example, a $1,000 monthly budget, historically driving five conversions a week, jumping to 15 per week is amazing (300% increase)!
A $10,000 monthly budget driving 300 conversions per week jumping to 400 is equally amazing (33% increase).
Yet if we only looked at the percentages, we’d assume the $1000 budget far outperformed the $10,000 budget.
This is why percentage improvements need to be grounded in historical context and ideally leverage enough data to be statistically significant.
You can use the compare function in the ad network to manage grids and reports.
Measuring Success When You Don’t Have Conversions
While many of the core concepts will still apply here – there is no denying you’ll have to adapt your measurement style to account for lack of ad network conversions.
This likely means you will be using manual bidding or data acquisition bidding strategies (smart bidding requires conversions to function).
Metrics at Your Disposal
This will be one of your most important search metrics, because it paints a picture of what the market looks like and how much you’re capturing.
Average CPC to Query Quality
While average CPC is important on the whole, it becomes slightly more important when there’s no conversion data.
Understanding which queries can be secured for which price is an important step in ensuring budgets are being directed to profitable prospects.
Engagement rate/Earned Views
Seeing which content is eliciting your prospects to not only engage but share your content is an invaluable success measure.
It helps you understand if your brand is “landing” or if it’s noise folks are ignoring.
I love click-type segmentation because it helps me troubleshoot whether campaigns are truly failing or if we’re just not tracking correctly.
Specifically, seeing “real world” interactions and call actions are helpful when guiding a client who might not be able to track conventional conversions.
Measuring success in PPC campaigns has to be a data-oriented exercise, but it also needs to factor in adaptive thinking.
Never abandon a channel just because you can’t track “normal” conversions.
You’ll just need to be extra vigilant about your query, click, and engagement quality.