Re: CPM Pricing Models

From: Bill Henning <bhenning_at_phoenix.door2net.com>
Date: Fri 18 Jan 2002 08:15:29 -0500

I've been seeing a lot of advocacy of CPA models, and I
felt it was time (again) for a small/medium-ish web
content providers point of view.

Rob, please understand I am not picking at you - and I am
delighted that CPA works for you; but as I recall, you
have a product to sell... which may have the tendancy of
causing you to prefer a CPA model.

Rob Frankel <Rob_at_RobFrankel.com> wrote:

> This is going to be a firestorm-starter, as CPM vs. CPA is a big
> issue. Although there are merits to both, my feeling is that now
> more than ever, advertisers want to see return on investment that's
> trackable. In my book, that means they're looking for CPA (cost per
> action). The days of the "promise of the internet", where billion of
> eyeballs flood through your home page are pretty much over.

I appreciate why advertisers would want CPA!

Allow me to give you a simplified example.

Say Acme Widget Co. talks Acme Journal into running ads
on a CPA basis.

Let's play with the numbers a bit.

Here is a table for estimated clicks based on a CTR percentate
and number of page views:

        0.20% 0.40% 0.60% 0.80% 1%
200000 400 800 1200 1600 2000
400000 800 1600 2400 3200 4000
600000 1200 2400 3600 4800 6000
800000 1600 3200 4800 6400 8000
1000000 2000 4000 6000 8000 10000
1200000 2400 4800 7200 9600 12000
1400000 2800 5600 8400 11200 14000
1600000 3200 6400 9600 12800 16000
1800000 3600 7200 10800 14400 18000
2000000 4000 8000 12000 16000 20000

Now I am sure some will argue that with a perfectly matched
audience and well written text links you can get 4% to 6%
CTR rates - fine, so multiply my 1% figures with whatever
you feel is appropriate.

Here is a table for estimated sales per click, based on sales
conversion rates shown:

        0.20% 0.40% 0.60% 0.80% 1%
400 0.8 1.6 2.4 3.2 4
800 1.6 3.2 4.8 6.4 8
1200 2.4 4.8 7.2 9.6 12
1600 3.2 6.4 9.6 12.8 16
2000 4 8 12 16 20
2400 4.8 9.6 14.4 19.2 24
2800 5.6 11.2 16.8 22.4 28
3200 6.4 12.8 19.2 25.6 32
3600 7.2 14.4 21.6 28.8 36
4000 8 16 24 32 40
4400 8.8 17.6 26.4 35.2 44
4800 9.6 19.2 28.8 38.4 48
5200 10.4 20.8 31.2 41.6 52
5600 11.2 22.4 33.6 44.8 56
6000 12 24 36 48 60
6400 12.8 25.6 38.4 51.2 64
6800 13.6 27.2 40.8 54.4 68
7200 14.4 28.8 43.2 57.6 72
7600 15.2 30.4 45.6 60.8 76
8000 16 32 48 64 80
8400 16.8 33.6 50.4 67.2 84
8800 17.6 35.2 52.8 70.4 88
9200 18.4 36.8 55.2 73.6 92
9600 19.2 38.4 57.6 76.8 96
10000 20 40 60 80 100

So, a site after 1,000,000 impressions with a 1%
CTR and a 1% post click sales conversion rate will
can be expected to net 100 sales.

Say the widget costs $50 (plus $10 s/h) and that the
merchant will pay 5% of gross sales.. that is $2.50
per item sold.

The site providing 1M impressions will gross $250
for running 1M ads, assuming that the merchant is
honest and reports ALL sales - or is even technically
capable of tracking them!

$250 / 1M = $0.25CPM

> There are still those who argue that CPM can work. It may, to some
> extent. But the main ingredient is huge volumes of traffic, so that
> the industry average of .25% clickthrough can produce anything
> meaningful.

When the merchants offer creatives with URLs plastered
over them, and then offer 2% of sales (an actual offer from
a well established web tech retailer to my tech site) you
KNOW the publishers are being taken advantage of.

Due to their branding creative I knew the CTR was going to be
about 0.4% (I ran some CPM campaigns for them thru some
brokers earlier at an OK CPM before the .BOMB).

Say I dedicated 1M p/v of inventory to them... referring to
the tables above, even assuming a 5% sales conversion it
would generate 200 directly trackable sales; average sale
was $50 - that $1/sale, or a whopping $200;
resulting in $0.20CPM

That won't even pay for my co-location.

I'll bet that a lot of readers - if they were interested in
the offer - would just write down the easy to remember URL
and visit it later, with no sales credit to me.

It would also stick in the minds of my readers (branding).

> Personally, I prefer lower volume, high-quality traffic tied to a
> clickthrough. I've found it delivers much high clickthroughs (4%)
> and happier clients.

That's great, if you find the right publisher match, and
offer a decent payout, and excellent (and fair) tracking.
Oh yeah - and you don't try to entice the reader to make
the order off-line with URLs, 800#'s... and have action
messages on the ads and on the site.

What I find especially frustrating is that some forms of
CPA could work... as could CPC w/o URL's, 800#'s and an
action message - but wait, that would mean
that the advertisers would actually have to pay for
advertising!

I have NO idea why advertisers are willing to pay $5-$20
CPM for untargeted eyeballs on Yahoo! yet are not willing
to pay $5 on highly targeted sites. No way the conversion
rate could be better at Yahoo.

In my experience, CPA campaigns are usually quite
specifically designed for branding effects, with minimal
payment to the publisher being a specific goal.

CPM campaigns are designed for maximal CTR rates, again
with minimal payment to publishers.

What does this mean for everyone?

Less content. I've had to concentrate on my consulting
business; and I am only now going to start reviving my
tech site - but until the ad business picks up I will not
be able to publish at the frantic pace I was doing when
ad money was available. I have a mortage and car loan to
pay. Other sites were not so lucky; as nearly as I can
tell a good 60% to 70% of tech sites have failed over the
last year - they did not have consulting to fall back onto.

Sure, you may say I am a "small fry" publisher - and you
are welcome to your opinion - but my readers who value my
advice, and the help they have gotten from my articles
would disagree with you. I truly wish I could dedicate
more time to running my site - I found it an immensly
rewarding experience, and I have saved all the "Thank you"
letters. However the ad dollars are NOT there,
not unless I am willing to give up my editorial integrity
and write on spec - something I am not going to do.

I miss AdAuction; I did well with them.

I miss the heyday of FlyCast, now Engage... they still owe
me some.

I miss Burst - I left them when I did not like their new
agreement; and they were not getting me enough targeted
ads to matter.

I used to miss BottomDollar, but now it looks like
PriceGrabber will work.

I used several other agencies, but most were less
than spectacular.

I hate popups, but I made decent money running About.com
popus. Lost a lot of readers (about 10%) and got a lot of
hate mail.

Ah well, enough venting for now.

Best Regards,

Bill Henning
Editor, CPUReview.com



Received on Fri Jan 18 2002 - 07:15:29 CST