Latest / The Yield Doctor with James Deaker / Retail Media Networks Grow in Predictable Stages — and Each One Has a Trap

Transcript

Retail Media Networks Grow in Predictable Stages — and Each One Has a Trap · The Yield Doctor with James Deaker

• 0:00
  Most retail and commerce media networks
• 0:04
  don't fail because demand disappears.
• 0:07
  Instead, they stall because they are
• 0:09
  built for one stage of growth and then
• 0:12
  the business changes underneath them. In
• 0:16
  2010, I created a path to excellence
• 0:19
  framework for publishers who wanted to
• 0:22
  build better pricing and yield
• 0:24
  management teams. And what I learned was
• 0:27
  that growth doesn't happen smoothly. It
• 0:30
  happens in predictable jumps followed by
• 0:33
  very predictable bottlenecks.
• 0:35
  And the same pattern shows up even more
• 0:38
  clearly today in retail and commerce
• 0:40
  media networks. So I've updated that
• 0:43
  framework. And while my focus today is
• 0:46
  on retail media networks because they
• 0:48
  are one of the biggest growth areas in
• 0:50
  our industry, the same ideas are equally
• 0:53
  going to apply to other fast growing
• 0:55
  publishers, whether they're CTV, gaming,
• 0:59
  marketplaces, or something else. You're
• 1:02
  going to find that you hit many of the
• 1:04
  same changes and challenges in a similar
• 1:07
  order. Teams launch quickly, revenue
• 1:10
  looks promising, and then suddenly sales
• 1:12
  starts to slow. Advertisers ask harder
• 1:15
  questions, operational cracks begin to
• 1:18
  show, and growth hits a ceiling that
• 1:21
  nobody quite owns. So, the mistake that
• 1:25
  many teams make is assuming that they
• 1:27
  either need more sales or more
• 1:30
  technology when the real issue is that
• 1:32
  the organization hasn't evolved to the
• 1:35
  next stage of the business.
• 1:37
  Today I'm going to walk you through a
• 1:39
  simple framework that explains how the
• 1:41
  retail media networks actually grow, the
• 1:44
  capabilities required at each stage, and
• 1:47
  why the right decisions early on make
• 1:50
  everything easier later. As you can see
• 1:54
  in this graphic, there are five clear
• 1:56
  stages that an organization goes
• 1:58
  through. I have deliberately portrayed
• 2:01
  the fact that they overlap and they do
• 2:04
  and you could argue that the challenges
• 2:06
  of the last two stages should be flipped
• 2:09
  depending on the organization.
• 2:11
  But let's get into it. As you can see in
• 2:14
  this graphic, the first stage is
• 2:16
  launching. It's from zero to ads. It's
• 2:19
  everything around what you need to do to
• 2:22
  get ads up on your page. The second is
• 2:26
  validating which I also call proving
• 2:28
  that you belong and it's the first year
• 2:30
  of operations.
• 2:32
  Beyond that we get into optimization
• 2:35
  which is making it all work. And then
• 2:38
  the last two stages which as I say could
• 2:40
  be interchangeable are called expand
• 2:43
  where you start to hit a growth ceiling
• 2:45
  around often year two or three. And the
• 2:48
  fifth is scaling with data which is
• 2:52
  taking control of those data assets.
• 2:54
  again, usually about 3 years out. So
• 2:58
  with that in mind, let's start at the
• 3:00
  beginning. Stage one is launching. And
• 3:03
  this is the act of going from nothing on
• 3:06
  your pages to having ads up on your
• 3:09
  website, selling those ads, and having
• 3:12
  the mainline business acknowledge that
• 3:15
  there really is a companion business
• 3:17
  that now exists around advertising.
• 3:21
  While there is a path where you could
• 3:23
  build this business using solely
• 3:25
  programmatic advertising hooking up to
• 3:28
  Google Adex, in practice, most
• 3:31
  businesses find that a better path is to
• 3:34
  invest in a sales team with a good sales
• 3:37
  leader. You also need some engineers to
• 3:40
  put ad blocks on the pages and ensure
• 3:43
  that you are delivering against them.
• 3:45
  So, what does it look like? Again, the
• 3:48
  main focus is on sales and ad product.
• 3:51
  You need to have a sales team that's
• 3:53
  either selling on a sponsorship or a
• 3:56
  flat fee basis or maybe they're selling
• 3:58
  some CPMbased products. And the key here
• 4:02
  is that initially they're going to go
• 4:05
  out to existing partners, sometimes
• 4:07
  referred to as endemic advertisers.
• 4:10
  These are the partners that the retail
• 4:13
  or the commerce business already works
• 4:15
  with who already understands the
• 4:17
  business and asks them if they are
• 4:20
  interested in some advertising space. So
• 4:23
  at this stage what are the key
• 4:25
  challenges? The key challenge is simply
• 4:27
  to try to get partners or advertisers to
• 4:30
  try the new retail media network to try
• 4:33
  to prove out that value to the
• 4:36
  advertiser.
• 4:37
  And there is a short list of things that
• 4:40
  you absolutely need to get right to
• 4:42
  build that foundation. The first is
• 4:45
  hiring. You need to have good
• 4:48
  salespeople and particularly a good
• 4:50
  sales leader and you need to have good
• 4:53
  product and engineering people who can
• 4:55
  ensure that things are running
• 4:56
  seamlessly.
• 4:58
  The second thing is good ad units.
• 5:02
  Meaning, you know, this sounds easy, but
• 5:04
  you need to make the decision. Are you
• 5:06
  putting native ads on your page? Are you
• 5:09
  putting traditional graphical banners?
• 5:11
  Are you going to focus on video or
• 5:13
  something else? And you need to ensure
• 5:15
  that these ads are viewable. So, one of
• 5:18
  those things that sounds easy, put a
• 5:20
  slot on a page, but a lot of decision
• 5:22
  making needs to go in because some of
• 5:24
  those decisions are going to come back
• 5:27
  and stay with you for many years.
• 5:30
  The third thing is you really need a
• 5:32
  scalable tech stack. I've seen small
• 5:35
  retail media networks who pick a niche
• 5:38
  technology because it's cheap and
• 5:40
  they're not sure how big the business is
• 5:42
  going to grow. And generally this is a
• 5:45
  mistake. You need to build a technology
• 5:48
  stack as if you are going to be
• 5:49
  successful later because it is extremely
• 5:52
  difficult to swap out an ad server to
• 5:55
  swap out your order management system if
• 5:58
  at all possible. You need to pick
• 6:00
  something early that is going to last
• 6:02
  you the whole way through this journey.
• 6:05
  And then the last component that you
• 6:08
  need to focus on is a rate card. And
• 6:10
  here it's the ad products and the
• 6:12
  pricing. And note that I said add
• 6:14
  product, not ad unit. I previously said
• 6:17
  ad unit or ad format, meaning what are
• 6:20
  the things that you're putting on the
• 6:22
  page. An ad product is a broader concept
• 6:25
  that incorporates both the ad format,
• 6:27
  but also how you're selling it. Are you
• 6:29
  selling it on a sponsorship basis? Are
• 6:32
  you selling on a CPM basis? What
• 6:34
  guarantees are you making around
• 6:36
  delivery? What guarantees are you making
• 6:38
  around viewability, etc. I've done a
• 6:41
  host of different videos on this which
• 6:43
  I'll I'll link to. So, I'm not going to
• 6:44
  blavorver the point here. Do you need a
• 6:48
  yield management team? At this stage,
• 6:50
  you probably don't need a yield
• 6:52
  management team, but you do need to have
• 6:55
  a couple of crucial decisions done that
• 6:58
  often might get done by somebody with
• 7:00
  yield management expertise. And the
• 7:02
  first of those is the rate card. Because
• 7:04
  again, the rate card is the primary
• 7:06
  mechanism for describing what your
• 7:09
  products are and how they're going to be
• 7:12
  sold and how they're going to be used by
• 7:13
  your advertisers. That can be done by
• 7:16
  someone with experience in the industry.
• 7:18
  It doesn't necessarily need to be a
• 7:20
  yield team, but it's important that you
• 7:22
  get it right and early. All right, so
• 7:25
  that's stage one. Stage two, validation.
• 7:30
  This is proving that you belong. This is
• 7:32
  the stage where the ad network is
• 7:35
  launched and you are promoting it and
• 7:37
  everything feels new to both you and
• 7:40
  your partners. So what does it look
• 7:42
  like? Some partners are running large
• 7:45
  campaigns. There's likely not many
• 7:48
  advertisers running at the same time
• 7:51
  because you probably haven't sold to
• 7:52
  that many. [gasps] The processes are
• 7:55
  going to feel ad hoc and each campaign
• 7:58
  is going to feel like you are building
• 8:00
  something from scratch.
• 8:02
  So what are the key challenges? The
• 8:05
  partners are going to expect reporting
• 8:07
  and measurement at the same level that
• 8:10
  they would expect from other partners.
• 8:12
  And there are clear things that you need
• 8:13
  to do to deliver on those expectations.
• 8:16
  So what are the things that you need to
• 8:18
  get right? The key here is good
• 8:22
  operations. You need to prove that you
• 8:24
  can get ads up and running, rendering on
• 8:27
  your page, and you need to prove that
• 8:30
  they can deliver. At this stage, do you
• 8:33
  need a yield management team or a yield
• 8:35
  management person? No. Again, for
• 8:38
  similar reasons to stage one, you can
• 8:40
  probably get away without it so long as
• 8:43
  you have people who are experienced in
• 8:45
  the industry who can make some of the
• 8:47
  right decisions.
• 8:49
  All right, stage three. And again, I'm
• 8:52
  calling this optimization, calling it
• 8:54
  making it work. This is you're a year or
• 8:57
  two into the journey. You have the
• 8:59
  business up and running for a year and
• 9:01
  it is starting to feel sustainable, but
• 9:04
  you have some significant challenges.
• 9:06
  So, what's it look and feel like? It's
• 9:08
  starting to feel like a sustainable
• 9:10
  business. But one of the key challenges
• 9:13
  is you may still be getting
• 9:15
  experimentation and innovation budgets
• 9:18
  from agencies and advertisers. In other
• 9:21
  words, they're playing around, they're
• 9:23
  interested, they're running test
• 9:24
  campaigns, but you have not fully proven
• 9:27
  out the value. They're not running
• 9:30
  repeatable campaigns because you haven't
• 9:32
  necessarily pro proven out the rorowaz.
• 9:36
  So you're getting feedback from
• 9:38
  advertisers about the things that they
• 9:40
  like and don't like about the ad
• 9:42
  performance about the viewability etc.
• 9:46
  So at this stage what do you have to get
• 9:47
  right? You have to get good reporting
• 9:50
  running. You have to start optimizing
• 9:52
  your ad units maybe moving them around
• 9:54
  on pages changing the the size of those
• 9:58
  units. You need to start having approval
• 10:00
  processes on the deals that you're
• 10:02
  running because you've got more
• 10:03
  campaigns coming in. Key operations at
• 10:07
  this stage is still really important.
• 10:09
  You need to prove that the ads are up
• 10:11
  and running and they are delivering
• 10:13
  value. At this stage, do you need a
• 10:16
  yield management person or team? Maybe
• 10:20
  you're starting to take on some of the
• 10:22
  different components of an organization
• 10:25
  that a yield management team would do.
• 10:27
  business reporting, deal approvals, rate
• 10:31
  card approvals, inventory forecasting.
• 10:35
  You either need a yield management team
• 10:37
  or you need those components done by
• 10:40
  other teams like ad ops or ops finance
• 10:44
  and it may be easier at this stage to
• 10:46
  start investing in a yield management
• 10:48
  team. You certainly need to have clarity
• 10:51
  as to who's responsible for what. All
• 10:54
  right, stage four, expansion.
• 10:57
  For most retail and media and commerce
• 11:00
  networks, if they are successful,
• 11:02
  they're going to start hitting limits to
• 11:04
  the growth on their own site. Maybe they
• 11:07
  have run out of inventory uh or at least
• 11:10
  the inventory that advertisers demand on
• 11:13
  the front page or list alongside key
• 11:15
  products. And so the challenge becomes
• 11:19
  where do you find growth in other ways?
• 11:22
  Uh and so the key question is where do
• 11:24
  we go from here? You've reached the
• 11:27
  limits of the first version of your
• 11:29
  business and you need to start planning
• 11:32
  for the second. And again I fully
• 11:34
  acknowledge that it is possible to to
• 11:37
  launch a retail media network running
• 11:39
  both on your own site and off network at
• 11:43
  the same time almost from day one. But
• 11:46
  in practice at some stage the limits to
• 11:49
  your owned and operated capacity are
• 11:52
  going to have to get addressed. So here
• 11:55
  we are looking at where do we go to get
• 11:58
  more supply and more demand? And that's
• 12:00
  going to involve questions like do we
• 12:03
  need new geographies? Do we need new ad
• 12:06
  units? Do we need new managed
• 12:08
  advertisers who are for example
• 12:11
  non-endemic? They haven't been partners
• 12:13
  in the past. Do we need new ad products
• 12:17
  like audience extension, like bundles,
• 12:20
  etc.
• 12:21
  Should we be focused more on margin
• 12:24
  management? Again, in those first few
• 12:26
  stages, you were just focused on getting
• 12:29
  things up and running, maybe a little
• 12:31
  bit of optimization and ensuring that
• 12:34
  the advertisers are happy. You're not
• 12:36
  spending a lot of time on the ad network
• 12:38
  margin management.
• 12:40
  You're probably also looking at
• 12:42
  partnerships with other publishers or
• 12:44
  other retail media networks and you may
• 12:47
  be starting to maximize your
• 12:49
  programmatic channels by bringing on
• 12:50
  things like private marketplaces through
• 12:53
  deal IDs. Now, do you need a pricing in
• 12:56
  your management team? Yeah, probably.
• 12:59
  Similar to uh the previous stage, you
• 13:03
  can maybe get by without it so long as
• 13:06
  other teams are doing it. But now you're
• 13:08
  really starting to look at options and
• 13:10
  tradeoffs that don't clearly fall in
• 13:13
  anybody else's job description. This
• 13:16
  isn't a sales activity. It isn't a
• 13:18
  finance activity, and you need a group
• 13:20
  who's independent. In the short term,
• 13:23
  your ad product team may take the lead,
• 13:26
  but they're not going to be managing it
• 13:28
  dayto-day. And now, let's get to stage
• 13:31
  five, scaling with data and taking
• 13:33
  control. And again, this may be done in
• 13:36
  conjunction with stage four as part of
• 13:39
  where do you expand.
• 13:41
  So at this stage, you're really starting
• 13:44
  to take on more complexity of your
• 13:46
  organization.
• 13:47
  Um there would may be organizations who
• 13:50
  are smaller who start with data, but I
• 13:52
  really realistically don't think they
• 13:54
  have the resources to do it. What does
• 13:57
  it look like? We are starting to do
• 13:59
  things with data and technology that not
• 14:03
  all the people who started with us in
• 14:05
  stage one understand. The questions
• 14:07
  we're asking ourselves is how do we take
• 14:10
  control of our destiny and how do we
• 14:13
  ensure that our data assets are being
• 14:16
  well used. At this stage, the key
• 14:19
  challenge is the business is starting to
• 14:22
  become really complex and hard to manage
• 14:25
  and not everybody in the business
• 14:28
  understands all parts of the business.
• 14:31
  So you are starting to do things like
• 14:34
  explore building your own ad server in
• 14:37
  order to take better control of the
• 14:39
  priorities of the ad serving. You're
• 14:41
  starting to look at either selling your
• 14:44
  data or selling different targeting and
• 14:47
  segments to your advertisers. You're
• 14:49
  looking at augmentation through
• 14:52
  targeting and through the creation of
• 14:54
  data clean rooms. And you're starting to
• 14:57
  look at dynamic floor management. Again,
• 15:00
  this is dynamic floor management. I've
• 15:01
  done videos on this. People love to race
• 15:04
  to do this as one of the first yield
• 15:06
  management activities. I would bring it
• 15:08
  on at this stage as one of the last
• 15:10
  activities that I would be exploring to
• 15:13
  see if it has value. Do you need a
• 15:15
  pricing and yield management team? Yes.
• 15:18
  Again, whether you formally create one
• 15:20
  or not, you need a group who is doing
• 15:23
  stuff that wouldn't normally have been
• 15:25
  done in the organization in those first
• 15:27
  couple of stages. So, you need very
• 15:29
  clear rules. You need the ability to
• 15:32
  manage tradeoffs across the different
• 15:34
  product lines and different channels.
• 15:36
  and you need the discipline to ensure
• 15:38
  that nothing falls through the cracks.
• 15:41
  Uh these challenges are ones that retail
• 15:45
  media networks, commerce media networks
• 15:47
  face time and again. And if I had one
• 15:50
  lesson that uh to take away from all of
• 15:53
  this, it's that you have to ensure that
• 15:56
  you are both understanding what stage
• 15:58
  your business is currently in but also
• 16:01
  planning for the challenges of the next
• 16:03
  stage. These are the types of
• 16:06
  interesting problems I work with my
• 16:08
  customers and clients on on a regular
• 16:10
  basis. If you need help, please reach
• 16:13
  out. I am James Deika. I am the yield
• 16:15
  doctor. Thanks [music] for watching.