2 Optimisation Strategies: Which Do You Use? (13)
What’s the first thing that comes to mind when you think optimisation?
For many affiliates, it's this:
KILL, DELETE, REMOVE, SCRAP, BANISH FOREVERMORE
Whether it’s a budget issue, or a psychological block, there’s no doubt in my mind that a lot of affiliates are over-optimising to the point where progress is nearly impossible.
That’s what this post is about.
The Whitelister vs. The Blacklister
There are two very different optimisation strategies:
Whitelist optimisation: targeting only placements X, Y and Z.
Blacklist optimisation: targeting all placements except X, Y and Z.
Neither strategy is right or wrong — and both can prove their worth in different environments — but one strategy is infinitely more scalable than the other.
If you want to be a super flying King of the Mountain affiliate, I believe the blacklist approach is your best bet, and here’s why…
Let's say you've launched a RON banner campaign on Exoclick.
You leave it for 24 hours, then check back.
Surprise, surprise. Try not to capsize in your chair: it’s a fucking mess.
Of the 3000 placements, about 10 are profitable, 20 are break-even, 70 are definite losers, and 2900 don’t have enough data to be judged a success or failure.
Where do you start?
Whitelist Marketer Approach
He downsizes the campaign and focuses on his immediately profitable targets.
This is a popular tactic with affiliates on a tight budget who can’t afford to leak money. They don’t want to run at a loss, and they can see a section of the campaign that is profitable.
Temptation is a bitch...
So Whitelist Marketer groups the 10 profitable targets in to a new campaign, calling it something stupid, like, I don’t know… "DE 300s Cluster of Profit Forevermore".
He now cranks up his bid to scoop as much of the profitable traffic as he possibly can.
What’s wrong with this strategy?
At first glance, not much.
It's precisely the form of 'pocket of profit' targeting that I’ve advocated in the past.
The trouble is, assuming you have aspirations of being more than a Pocket Change Marketer, it whacks a giant glass ceiling over any hopes you might have of scaling it mahoosive.
Three reasons why:
1. Whitelisting is an instant volume killer. By defining a ‘whitelist’, you are committing to a handful of placements that worked yesterday. There’s no guarantee they’ll work tomorrow — and in most cases, the guarantee is void by daybreak.
2. Shutting out new targets means you won’t know if there is new lucrative inventory in your market — pretty much the reason for going RON in the first place.
3. Banner blindness bites twice as fast, twice as hard.
If you are dead set on taking the whitelist approach, I would strongly recommend focusing on the highest volume placements (that are still profitable) as individual campaigns per single site. Not in a cluster.
You then need to actively manage these campaigns on a day-to-day basis, replacing creatives like a madman, and remembering not to create too many similar campaigns lest you reach your executive capacity -- before it all goes tits-up fast.
The good thing about this whitelist strategy is that it enables affiliates on shoestring budgets to compete by outworking the competition.
The reality is that guys like myself who have been doing these campaigns for years don’t have the hunger to make a single RON placement work in the same way that a guy desperate for anything with a positive ROI might.
Blacklist Marketer Approach
The second strategy requires balls of steel and a budget capable of withstanding early blows.
The Blacklist Marketer doesn’t bother launching separate whitelist campaigns. He focuses on monetising the entire market — whilst weeding out the very worst performing placements.
He throws out the sprouts but keeps the meat and potatoes. If you're not British, that's going to sound fucking insane, so bear with me...
Instead of whitelisting the 10 profitable targets, he’ll blacklist the 70 definite losers and leave everything else running.
Instead of over-optimising, he’ll focus on creating a better overall funnel to improve ROI.
That means:
- Building a better landing page.
- Coming up with better ads.
- Finding a better offer* (This makes pretty much every other factor redundant if you haven't already found it)
- Improving the delivery of the content.
- Upselling or harvesting long-term value.
What the Blacklist Marketer won't do is look at yesterday's stats and pretend he knows exactly which 10 out of 3000 placements are right for his lead generation business, cull the rest, then try to live happily ever after.
Note: Some traffic sources are front-loaded with placements that make up a huge percentage of the entire RON traffic. Like TrafficForce and Porn.com. In these cases, you might as well just run a premium placement on Porn.com. You have no business setting up RON campaigns if you can mindmap and test the entire inventory in 20 minutes.
So, blacklisting is my preferred strategy for dealing with RON traffic. And it's what I'd recommend to any affiliate who cares about scaling hard and fast.
Three reasons:
1. You reap maximum volume, including the many targets with stats that look like this:
How are you going to whitelist profitable targets when you don’t even know they exist?
(Now that’s some chicken and egg shit for you.)
2. If lucrative inventory arrives on the network, I’ll get to it first with my RON campaign. The whitelist marketer is too busy putting out fires in his small cluster of once-upon-a-time-profitable placements to fully appreciate that tomorrow’s stats are up for grabs outside his bubble.
3. If you want statistical significance across 3000 targets, you're more likely to get it by culling the worst targets. Not by pinning hopes on the early leaders.
Blacklist or Whitelist?
I don't want to suggest that you have to launch vast RON campaigns without ever culling a target.
That's not true.
There are times, and indeed entire networks, where a whitelist strategy will make you more money.
But one of the big, big problems I see is an eagerness from affiliates to clutch at profitable placements -- often at the expense of improving their funnel and making it work across the board.
That, for me, is how you know that you have a big campaign on your hands: when the placement is no longer the issue.
So, what's your preference?
Are you a whitelister or a blacklister? Do you close your net around a small handful of placements, or do you blast it far and wide over any inventory you can get your hands on?
damn, finch you are a real mind blower, what i do is usually create a new campaign, whitelist the profitable target, then blacklist the definite loss target from original. keep both campaigns running.
the whitelist campaign i call winning-pool, the blacklist one i call testing-pool.
the headache of RON + blacklist is lots of target with very few data, especially if you run a RON CPM campaign, you are too hard to figure out the real CTR.
so volume means a lot to me while a RON campaigns running, you have to figure out the conversion target's CTR, to optimize it by creatives.
anyway, my option is if your cash-flow is not enough, whitelist campaign is a fine way to go, but it kills your volume indeed.
RON always hurt quality as well, most of the time the advertiser cut you out before you make it green.
in my experience, to run a blacklist campaign
1. you have 2-3 successful lander.
2.sort your losing target, cut the IDs that spent 2x times of your payout (most people suggest 3x times, but if you run a RON campaign, 2x is really enough)
3.sort your click volume by target, cut the IDs that dont click the lander or the lander CTR way too low.
4.sort your click volume by target, cut the IDs that click a lots but dont convert.
5.sort your your conversion volume by target, cut the IDs that have unusual EPC. (option to those countless targets campaign)
and if you want a whitelist campaign, here is what i do:
1.sort out the high volume target that convert, optimize them - whitelist campaign 1
2.sort out the high epc target. bid higher - whitelist campaign 2
just my 2 cents, anyone correct me please.
in the end, when will you drop the new premium post? i ve been wait for a lifetime man.
I'm with you, blacklist all the way.
Get the full spectrum and use the data to optimise while not hamstringing the campaign's legs.
Great post mate!
Is there such thing as a grey list strategy?
The more I master a particular traffic source or niche the more of an idea I have how a particular campaign will perform before running data to it.
On some traffic sources I will blacklist some placements that are always terrible performers, I will also target particular criteria because it worked well previously.
However, if I move to a new source running a new niche, I tend to always opt for the blacklist strategy.
Great post! Always go blacklist way myself.
What would be the sort of budget needed to do this approach? I know thats a "how longs a piece of string" type question but just so I can get a feel for this approach.
So for example in this sample with 3k sites would it be $1k a day, $2k, $5k?
Originally Posted by Mr Green
Originally Posted by garyp
Beyond that, it depends on what countries you're targeting, and how much traffic the platform has.
You can make life easier by going RON in a less competitive (and smaller) country first.
the whitelist campaign i call winning-pool, the blacklist one i call testing-pool.
Another point I'd raise about RON campaigns is that you have to be very, very careful using Smart CPMs.
If your maximum bid is 0.60, and your Smart CPM means you're only bidding 0.30 on a big volume placement XYZ, it can be difficult to track when another advertiser comes along and starts pushing up that price. Suddenly you're bidding 0.60 without actually changing your bid.
If placement XYZ went a long way towards your profitability, it can smoke the overall ROI pretty damn fast.
Anyone use a tier system? Grouping zones into high/low/medium performers and using different bids for each group.
In theory, you have a super optimized campaign, but when I tried this my overall ROI went down (Ad manager said there was a tech issue or something)
Personally, I like to blacklist/change my bids until I'm happy with my ROI/volume. If I still can't get the campaign profitable then I just whitelist the best placements.
Great post!
Personally I run both types of campaign, but if I'm whitelisting I generally do it right from the start. Choose a huge site / app that I know has more traffic than the M25 in rush hour, research it within an inch of its life, and design banners, angles and landers to target that site's demographics specifically.
That's given me some reasonably nice campaigns in the past, with decent longevity.
But if I'm starting out aiming to produce a general-appeal campaign across a multi-placement buy, then yeah, I'm all about the blacklist.
Great post.
I was doing the whitelist - pooling approach for a while and after few weeks it just die out and the exact same thing happened as Finch described. Now instead I'm trying the blacklist way - gather as much data as I can and just cull the biggest losers.
Will see how it will go - the only thing is - until I get somewhere - I had to change the main offer 2-3 times, so it's even more mess.
I used to do whitelisting, and it worked okay. The problem is when the market shifts.
What I'm doing now is finding the highest volume placements on a given network, and testing those placements on a per-geo basis. Looks like I'll be getting more results faster.
Home > Affiliate Marketing Forum >