Latest / The Yield Doctor with James Deaker / Why Revenue Can Fall Even When Every Product Raises Prices

Transcript

Why Revenue Can Fall Even When Every Product Raises Prices · The Yield Doctor with James Deaker

• 0:00
  One of the strangest meetings that you
• 0:03
  can ever get called into as a yield
• 0:05
  leader is something like this. Every
• 0:09
  line of business is reporting that their
• 0:11
  price realization has improved and yet
• 0:14
  the total company revenue is down. It
• 0:17
  sounds impossible, but this actually
• 0:19
  happens more than you would think in
• 0:21
  digital media. And today I'm going to
• 0:24
  walk through an example or a couple of
• 0:26
  examples of the problem that in revenue
• 0:29
  management and yield management is
• 0:31
  called the mix problem. So this week's
• 0:35
  video is based on something that I've
• 0:37
  seen many times over the years. And if
• 0:39
  it's not identified, it can go on to
• 0:42
  cause problems for a company trying to
• 0:44
  maximize its digital media yield. And
• 0:48
  what makes it tricky is that even
• 0:50
  well-intentioned teams can be doing
• 0:53
  exactly what management has been asking
• 0:55
  them to do and still ending up with less
• 0:59
  revenue. So let's walk through a simple
• 1:01
  example. Imagine a publisher with two
• 1:04
  products. We're going to call product A
• 1:07
  premium inventory and product B standard
• 1:10
  runofight inventory. Premium sells at
• 1:14
  higher prices, but both products are
• 1:16
  good products. All right, so let's
• 1:18
  imagine that this publisher has 200
• 1:21
  million impressions to sell. Premium
• 1:24
  sells at twice the price of the standard
• 1:26
  inventory, so it's selling at $20 CPM.
• 1:29
  The standard run of site is selling at
• 1:32
  $10 CPM. And we have $100,000
• 1:35
  impressions being sold against each. So
• 1:38
  premiums making $2 million in revenue.
• 1:41
  The standard run of site is making1
• 1:44
  million and the total revenue for the
• 1:46
  site is $3 million. Hopefully pretty
• 1:50
  straightforward.
• 1:51
  Now the management team gives the
• 1:56
  broader team a very reasonable
• 1:58
  instruction to improve price
• 2:00
  realization.
• 2:01
  So the teams responsible for both
• 2:03
  premium and standard go out and do that.
• 2:07
  The premium team in increases the CPM by
• 2:10
  10%. The standard run of sight increases
• 2:14
  by 10% also. And on paper everything
• 2:17
  looks like a success. But something
• 2:20
  subtle happens because premium has
• 2:22
  become slightly more expensive for some
• 2:25
  buyers and those buyers may begin to
• 2:28
  shift towards the standard product
• 2:30
  instead. And so before I show you the
• 2:33
  result, let me ask you a quick question.
• 2:36
  If both products increase their price,
• 2:38
  what do you expect to happen to total
• 2:40
  revenue? And most people would say that
• 2:43
  the total revenue should increase. And
• 2:46
  let's see what actually happened after
• 2:48
  the price change. The premium homepage
• 2:51
  has indeed increased its price up to
• 2:53
  $22.
• 2:55
  Standards up to $11. premium selling
• 2:58
  60,000 impressions and the standard run
• 3:00
  of site is selling all the rest of the
• 3:03
  site out 140,000 impressions.
• 3:07
  But now if we add up the revenue totals,
• 3:10
  what we see is that the total revenue
• 3:12
  comes to $2.86 million. In other words,
• 3:16
  slightly less than we were making
• 3:18
  before. And again, let's look carefully
• 3:21
  at what happened. Both products
• 3:23
  increased their price realization but
• 3:25
  the total went down. Why? Because the
• 3:28
  mix of products changed. More demand
• 3:31
  moved towards the lower valued product
• 3:34
  which resulted in lower overall yield
• 3:37
  for the business. And this is obviously
• 3:40
  a toy example and the answer is obvious.
• 3:42
  But in a real publisher, you can have
• 3:45
  dozens of overlapping products. You can
• 3:48
  have multiple sales teams, different
• 3:51
  pricing structures, and sometimes even
• 3:54
  different lines of business. And when
• 3:56
  this mix occurs, it can be extremely
• 3:59
  difficult to detect. And each team may
• 4:02
  honestly be reporting, "Our price
• 4:04
  realization improved." And they could
• 4:06
  all be correct, but the overall system
• 4:09
  for yield deteriorates.
• 4:12
  Now, before you say, "James, why are you
• 4:14
  wasting my time with such a simple
• 4:16
  example?" Let's change the problem just
• 4:19
  a little bit. Let's go back to the
• 4:21
  original setup. We start with two
• 4:23
  products. Premium at $20 and standard at
• 4:26
  $10. And we sold 100 million impressions
• 4:29
  of each, $3 million of revenue. Now, we
• 4:32
  introduce a new product, something very
• 4:35
  common in digital media and in retail
• 4:37
  media. We launch a new product,
• 4:39
  something called audience extension that
• 4:41
  allows us to sell impressions not just
• 4:44
  on our own site, but across the web. So
• 4:47
  instead of selling 200 million
• 4:49
  impressions on our own site, we now sell
• 4:52
  50 million on other sites for a total of
• 4:54
  250 million impressions. And at first
• 4:58
  glance, that sounds like great news. We
• 5:00
  got more scale, more impressions, and
• 5:03
  more revenue opportunity. But before we
• 5:06
  look at the the numbers, let me ask you
• 5:08
  a question. And you can probably guess
• 5:10
  where this is going. If the company
• 5:12
  launches a new product and sells 25%
• 5:16
  more impressions, what would you expect
• 5:18
  to happen to revenue? And most people
• 5:20
  should expect that the revenue would
• 5:22
  increase. But let's see what actually
• 5:25
  happens. The premium product now is
• 5:28
  selling less impressions, 40 million
• 5:30
  impressions. The standard run of site is
• 5:33
  selling 110. And we're selling audience
• 5:36
  extension
• 5:38
  both on network and offn network. So
• 5:42
  it's all classified as the audience
• 5:44
  extension product. And the revenue from
• 5:47
  that is significantly less. The revenue
• 5:50
  in this case is only going to be $2.7
• 5:53
  million. And so let's think about what
• 5:56
  just happened. The company launched a
• 5:58
  new product. It increased the total
• 6:01
  supply from 200 to 250 million
• 6:03
  impressions, created more scale and more
• 6:06
  activity, and yet the revenue fell from
• 6:08
  3 million to $2.7 million.
• 6:12
  Why? Again, the mix shift occurred.
• 6:15
  Audience extension didn't necessarily or
• 6:18
  didn't steal demand just from standard
• 6:20
  and inventory. It also stole it from
• 6:21
  premium, which is what can happen in
• 6:24
  real life advertising businesses. Buyers
• 6:27
  still want the audience. They just want
• 6:29
  to find it in a cheaper way. And that is
• 6:32
  exactly why the whole function of yield
• 6:34
  management exists. Somebody needs to be
• 6:37
  responsible for the system level
• 6:40
  economics of the business, not just the
• 6:42
  individual products because optimizing
• 6:45
  each product individually does not
• 6:47
  optimize the business as a whole. That
• 6:50
  oversight might sit with a yield
• 6:52
  management team, revenue operations, or
• 6:55
  sometimes finance, but someone needs to
• 6:58
  be watching the mix. So, let me be
• 7:00
  incred incredibly clear about the
• 7:03
  purposes of these examples. I am not
• 7:06
  saying don't try and increase the price
• 7:08
  realization of product lines.
• 7:10
  Absolutely, you should. And I'm
• 7:12
  definitely not saying don't launch
• 7:14
  launch audience extension. It can be an
• 7:17
  incredibly valuable product. The key is
• 7:20
  to make sure that you are not
• 7:22
  inadvertently cannibalizing your own
• 7:24
  demand. I am James Deer. I am the yield
• 7:27
  doctor. Thanks for watching.