AdExchanger had a nice post from TRAFFIQ’s Scott Portugal about
publishers using a Value-per-Customer metric for site visitors. As longtime readers know, I’ve been advocating publishers take this approach
and measure site performance on revenue per visitor for years – most recently
I especially liked this part of Scott’s post:
“users interaction with a site leaves even more valuable
breadcrumbs that should be used to target that user. An example data set for an
online content provider will likely include: average time on site, average
pages consumed, bookmarked landing page, links shared, number of comments left,
path of visitation throughout the site….and that's before we get into ads
clicked, registration data (which will likely include basic demography), ads
viewed per session, etc.”
I’m not sure I agree with Scott on the value of those
specific metrics to advertisers, but certainly I agree with segmenting and
targeting based on site actions and events and then understanding the revenue
derived from those segments at the visitor level in order to valuate each
visitor, page and action.
For smart advertisers using DSPs (Demand Side Platforms)
this is already happening. The “value-per” or “cost-per” idea has emerged to be
the dominant growth force on the demand side. Unbundling of impressions is the
backbone of RTB (Real-time-bidding). Instead of buying a block of inventory
through a network or exchange that is not well matched to an ad message,
impressions can be more efficiently targeted and purchased via cookie matching.
This is a once a decade opportunity for publishers to take
advantage of. It might be their last. Advertisers would love publishers to be
able to directly deliver matching data (it originates from publishers
already but is collected by third-parties). Believe me, advertisers are not married to demand side data collection
and use. Advertisers are agnostic to where they get data. Their concerns are
related to efficient spends and ad performance. Both will be markedly improved
with more robust publisher data and segmentation.
Data Crunch: Supply vs
The one advantage demand side valuation and data tools have
is scale. As my friend and web
analytics luminary Judah Phillips tweeted to me yesterday “cost-per” metrics are more effective business measures than gross tonnage.” Scale is required for platforms
but platforms are best dumb. As I alluded to above and in this post demand side data is never going to be better at
matching than supply side because it removes context and timing (not to mention privacy issues).
Advertisers excitement about RTB is putting market pressure on the supply side to valuate visitors and
content. For years CPM has at best been arbitrarily determined and at worst determined
in an opaque manner by third parties — precisely because publishers have
had no idea about the value of sites or pages. How crazy is it that the
advertisers know more about the value of the publisher audience than the
publishers themselves? This is not a good business position to hold in an
auction-based environment. Really understanding the value – better than
advertisers – is required (case in point, Google).
It’s incredibly obvious that all publisher visitors are not
worth the same nor is a three-tiered system valuation even close to reality. Why
is a search referrer typically classified as remnant? A visitor with intent in
the middle of goal flow is surely one of the most valuable visitors to a site.
The idea of premium, secondary and remnant is flawed and holding back
publishers. These classifications are sure to vaporize over the next few years
as we truly begin to value the visitor.
Without this value-per data how can pubs knowledgably set-up
bid floors? Are they really going to allow a third-party with their own
self-interest to do this for them?
When publishers can valuate on “cost-per” the exchanges and
RTB will be their best friend and the market will determine the upside value of
their audience based on performance. Sites with great performance will see CPMs
beyond what they could imagine. Proof is backing in successful Search campaigns
to CPM (of course Google does this for you). As I’ve mentioned numerous times a
successful CPA campaign can routinely be in the $50-$100 CPM range. This is
because of relevance and matching – efficient impressions. Unbundled
Display doesn’t yet have the search like ability for a keyword
to unbundle a publisher's impressions in realtime but I can tell you that it's getting very, very close and we're insanely focused on this at Yieldbot. Having this level of understanding allows publishers to take control
over the monetization efforts of their site. Publishers that don't will bleed
to death, stabbing their belly with their own ignorance.