It’s intimidating to test ad networks, especially when there are so many avenues to explore and platforms that could take your money. As an advertiser, your options are so expansive that you’ll either stay complacent with a network that’s not doing too much damage – or bounce between ad networks too soon and too often in hopes that one of them will stick. Neither option is efficient, and both will run your budget into the ground.
But the truth is, there is a way to strategically test ad networks that provides you with insight and data without spending thousands of dollars. It’s not about finding the perfect ad network; it’s about creating a testing system that shows you what’s functioning and what’s not before you scale.
Know Your Baseline Before You Begin
The first step before testing is knowing your baseline. What are your expectations for success? What do you currently spend to acquire your customers? What’s the cost of acquiring a customer versus their lifetime value? How much can you spend before you’re digging yourself into a hole?
You should know these stats by heart. They’re not just useful metrics to have; they’re the basis for your entire subsequent strategy. Without knowledge of your current situation, then every ad network is either performing well for you or poorly, but you’re just speculating based on conversion performance alone.
For example, an ad network could deliver a click to your website for $0.50. Still, if $0.50 doesn’t translate into sustainable conversion and you can justify spending a higher CPA to achieve your goals without losing money, then the low bid doesn’t matter.
You need a clear expectation of your respective CPMs and expected conversions. What’s working well (or not at all) for you now is also the baseline through which every other ad network must beat (or match) to get more budget allocation going forward.
Use Your Budget Intentionally and Sparingly
This is where most people go wrong: they discover a new ad network and spend a large percentage of their budget before realizing it doesn’t perform miracles in the first week. Either they give up or they too stubbornly continue spending because they’ve dug themselves into too deep of a hole to see it otherwise.
Instead, use a set testing budget that you can reasonably afford to lose for the sake of intentional and efficient testing without screwing up your established campaigns.
Your testing budget might range from $500-$2000 depending on the industry and your specific outreach, but it’s less about the actual number than it is about intentional discipline. When researching networks, advertisers generally look at the top ad networks available online to see what has the best targeting options and ad formats, so make sure that what you’re choosing will actually resonate with your target consumer as well as are the big named options.
For example, don’t pick 10 ad networks for testing; narrow it down to three-to-five at most. While it may seem easier to test all ten at once, if you’re only spending $150 each for statistical significance thresholds, then you’re better off testing three networks with $500-$700 each instead.
Find Common Variables To Test Across Ad Networks
When testing multiple networks for testing simultaneously, you want to use as many controlled variables as possible.
For example, use the same creative, offer in the copy, targeted demographic, bidding platform and one landing page that you’ll track with UTM parameters. You’re not testing what’s working on creative with what networks; you’re testing what’s driving the best quality traffic for what you need through network quality alone. If one network’s performing better with more robust CPI statistics than another’s and you’ve tested varied ads among different networks, you’ll never know if it was the network or just the creative acting poorly.
Also note that what you’re driving traffic to should be the same as well – a common landing page for submission – and track any conversions in a negative or positive direction via conversion tracking downriver to actual purchases completed, sign-ups submitted, or qualified leads generated based on your business model.
Give Every Network Enough Time
There’s nothing worse than giving up on a good thing too soon or throwing money away unnecessarily. When providing testing allocations to an ad network, give enough time or financial stability
Digital advertising doesn’t happen overnight; most networks need time to optimize delivery on their end, just as your campaigns need time to accrue sufficient data to paint a compelling picture.
At minimum run tests for 7-10 days – if you’ve spent enough of your allocation test budget to get 50-100 conversions (if that’s appropriate for your business model) in that time frame – great! If not, assess different success metrics related to micro-conversions since they may take longer in lower-volume spaces.
However, this doesn’t mean giving bad-performing networks additional time for sake of stability either; if engagement metrics are low and no conversions are occurring after getting 30-40% of the way through your budget allocation testing cycle for performance improvement – but you’ve seen no positive performance – it’s okay to pause and reallocate that sum elsewhere.
Track Metrics That Matter
Click through rates and impressions are pretty for digital reports, but unless they’re translating dollars and cents to bottom-line growth, who cares?
When comparing multiple ad networks, focus on metrics that financially benefit performance. Cost per acquisition is usually the number one metric that counts – what’s the cost per conversion attributed per network? Then look at which network yielded customers from what acquisitions in addition – are those customers spending more on average?
Some networks yield cheaper conversions; others higher-cost conversions. But if over time, you adjust attribution based on value proposition offerings after getting buyers in the door, then paying more to net higher lifetime values per transaction means nothing unless you’re tracking acquisition costs and longer term values associated with prospective customers.
Engagement metrics matter too; how long do people stay on your page? What’s the bounce rate? Are people engaging by looking at more than one page or navigating away quickly? If so, are they real traffic engaged in something good for your business – and once again – bringing up those conversion rates?
Analyze The Data Without Emotion
After you’ve run the testing cycle, take a beat before exploring life outside of a report. Too often people’s hearts steer them away from sound decision-making because they appreciate certain aesthetics that worked well for them before – even though they’ve got no positive performance numbers to show for it.
Don’t discredit good performance if its numbers don’t lie; create an easily digestible side-by-side spreadsheet showing metrics for comparison.
The most important stats will include total spend per network, number of conversions attributed per network, cost per conversion via CPA, average order value compared to conversions for ROAs as well as any other numbers critical for operational realities.
Spot clear winners and losers easily – not every network works for every brand. If one has lower quality metrics than baseline projections at 40%, give it credit – if another network costs twice as much but presents worse quality – ditch it without question.
It’s the results in the middle that make sense with further assessment – the arbitrage cases where it’s close – but not quite as good – as is worth exploring down the line – for now – but with tweaks made.
Read More: 6 B2B Lead Finder Tools You Can Use to Generate Leads
Scale What Works & Kill What Doesn’t
Once you’ve tested various networks and found some reasonably decent results (or not), move along with sacrifice criteria or ones with good cumulative findings or comments.
It’s great to scale what’s working – but don’t go from $500-tested network options up to $5000 in one day; trickle it up over time, 20-30% over subsequent weeks/two week intervals without upsetting operating expectations.
For networks that didn’t work well, don’t feel bad – for whatever reason – and sorry it’s a bad fit – but if it’s not working for your brand now – even if it’s a big name brand that people want to work with – don’t compromise success for whatever’s best for them.
Some networks are niche-focused; if they’ve been explored and it’s clearly not a good fit now – and may not be in the future either – better cut ties quicker than they start leeching revenue.
The mediocre ones – hold onto them until later with different creative tweaks or even new target markets – that maybe it could work before putting down – but not now unless there’s clear winner/loser ideology established.
Learning how to test multiple ad networks effectively does not have to burn through countless budgets experimentally; instead, there’s a systematic way of approach that champions causation through correlation without giving carelessly without stocking up new variables along with projections – with discipline tied to monitoring success proven effort goes a long way in determining what could really matter for your business without spending excess time figuring it out.
