Two conversations I've had in the last week have caused me to raise an eyebrow.
The first with with a colleague, about a treatment center we were considering working with.
"Every week," my colleague said, "they like to throw some money at some marketers, and see who produces results."
"How long do they give them?" I asked.
"A month," he replied.
In the second conversation, another colleague told me a startling detail: certain treatment centers are renegotiating their marketing contracts week-to-week.
And I get it.
Everybody's been burned, blah blah.
And testing marketing channels in parallel is actually a good way to determine what's working. A lot better than taking somebody's word-for-it.
There's just one problem: marketing channels take longer than that before you know if they're working or not.
First, a word on how the average internet campaign works, and the minimum length of time for which you should probably test it.
Second, some suggestions about how you could still remain cautious, and test everything, without cutting-off-your-nose-to-spite-your-face.
Finally, an examination of the "side-effects" when you give marketing channels so little time to work.
Climbing to Altitude
Radio and TV campaigns either produce-calls-or-they-don't. You buy time on local or national media, you run the ads, and, if you've selected your audiences carefully and produced a good ad, the calls some in. Immediately.
But most channels aren't like that.
Referrals aren't. You don't set-up-shop without knowing any of the primary care providers in your area and expect referral traffic to come trouping-in-the-door.
SEO sure isn't. You need months-to-years to start ranking for your keywords and seeing results.
The problem with paid internet channels like Adwords and our specialty, Facebook, is some treatment marketers assume they're going to either work or not work in short order, like radio or TV. They won't.
With radio and TV, you have to pay up-front for the media buy. You've got an audience that's performed well in the past, and you show them the ad. Presuming you hit-the-target, the calls come in.
With Adwords and Facebook, you don't know at the beginning who or where your audience is.
Sure, you could run ads to the entire US on Facebook, or target the most general, most expensive keywords on Google. But you'd spend way more money on ads than you'd earn back on admissions.
With Adwords, the magic lies in finding low-competition keywords you can rank for, so you can afford them, that also select for likely clients, so that ad spend isn't wasted. (That's getting harder-and-harder, because keywords are in an auction format, with the ad view going to the highest bidder, and, thanks to sites like Spyfu, anyone can see anyone else's keywords. That's why we prefer Facebook.)
With Facebook, it's all about finding the best audiences. What's the small pocket of people with an interest-in-common, who are likely avatars for your clients.
On both platforms, there's no way to guess perfectly what's going to work, so you do research, test a bunch of things in parallel, and iterate until you find a couple of "winners".
That's the human intelligence.
On Facebook, there's also AI. Facebook sees who's converting from your ads, and adjusts its targeting so it's showing ads to the people most-likely-to-be-interested.
Here's the thing about both processes: they take time.
The Facebook algorithm takes about two weeks to "learn" enough to lower the cost-per-action as low as it's going to go.
And the human optimization tests multiple audiences until one-or-two hit, or drop below, the desired acquisition cost.
So - how long does it take a paid internet ad to reach full steam?
Put it this way: in the first month of a campaign, we typically see cost-per-qualified-lead over $1000.
In the second month, it's below $750, and in weeks 5-and-on, it's down to $550.
We've seen clients get admits in the second week of a campaign, but it's not typical.
All-said-and-done, we recommend 90 days to evaluate whether an internet campaign is working, though you'll have some precursors:
Even if you haven't gotten any admits, is the cost-per-insurance-qualified-lead dropping below $600? If not, and it's been more than 6 weeks, that's a bad sign.
Are you getting qualified-insurance leads at all? If you're not seeing at least a few by week 2, that's also atypical.
But that's not what many marketing managers do.
Here's the more typical process:
- Launch a test campaign
- Observe that it's been 4 weeks, and admits are still-at-zero (because that's the only statistic that's legible to you), and assume the campaign's a failure, and shut-it-down.
As I said, we've seen some great early results, but those are outliers. What we do see, though, is the already great cost-per-qualified-lead beginning to generate great results around weeks 8-12. Imagine if you'd shut that campaign down after the first month.
By misunderstanding how real internet campaigns work, treatment marketers are creating some bad side-effects too:
Know who will guarantee you immediate "call volume" from week-one?
Somebody who isn't curating for quality at all.
Ever purchased a "block of calls" - let's say 100 - for, say, $50-a-call? Let me ask: how many of those turned into admits.
I explained in an earlier article what type of campaign I'd run if I just wanted to produce "call volume", agnostic to quality. But it's not ad-dollars-well-spent.
What to Do Instead
We encourage centers to test campaigns in parallel. (Our system can run in parallel to a center's existing social ads.)
But run each for at least 90 days.
Oh no! That's a longer commitment! How can I be reasonably sure I'm not going to waste 90-days-worth-of-budget?
Simple: You're going to have to educate yourself enough to tell a real marketing strategy from BS. Presenting: my article.
Once those campaigns are running, you also don't have to take the full 90 days.
Remember, if you're not getting some qualified leads in the first few weeks, that's a bad sign.
...and if your cost-per-qualified-lead hasn't dropped below $700 after 6 weeks, that's another bad sign.
So why use internet ads at all?
Paid internet advertising (and we think Facebook is the best channel) has a few advantages over other channels:
Unlike radio and TV, internet ads don't require a $50-70k budget just to get started, and they also won't "burn out" the same way a TV spot will. With small tweaks and good management, they'll continue to produce leads in perpetuity.
In our experience (and colleagues across the industry have echoed this), the cheapest qualified leads are on Facebook.
Finally, though internet ads aren't a perfect "plug-and-play", because you're still dealing with human beings, they're far more scalable and controllable than either SEO or referrals.
In my opinion, the centers who take the time to learn what really works in marketing, and have the patience and discipline to build an asset, are going to win, especially when non-branded advertising goes away.