In my last couple of posts I’ve written about how most of the money that’s being spent on advertising in digital media is a waste. Let me give you a simple way of assessing where your problems might lie and how serious they might be.
There are really just two things to look at:
1. What KPI you're driving and who controls measurement
2. How it’s being delivered - what media are you spending on and whom you’re spending it through
Here is a simple way of looking at it.
Score of 2: Very Healthy – there may be some opportunities to optimize but your budget is in your control and working for your business
Score of 5-8: Health Risk – Either budget or KPI measurement is not in your control
Score of 10: Immediate and high risk – you might be wasting your digital budget
(Originally posted on Linked In on Feb 20, 2016)
In some versions of the old story about the youth who discovers the hoarded treasure of a gang of thieves, each of the 40 thieves in the band was a specialist in a specific kind of crime. In a similar vein, there are several ways in which your digital budget is being mis-spent by various parties, and while I'm not going to go through a list of 40 here, let's look at some of the key ones.
The context here is paid media - where you pay partners to deliver a campaign using some form of digital advertising material - as opposed to agencies managing social media or a community for your brand.
As a basic setup to examine this topic, let's say you're a client looking for a non-purchase KPI - something as simple as getting people to view at least 15 seconds of a 30 second video. Let's say your target is to get 100,000 views in a specific demographic and geography.
Now, let's look at some of the big drains on your budget
Agency rebate : 10-35% of your budget could be disappearing into the agency's pocket merely because they're able to twist the arms of DSPs, ad networks and publishers, many of whom are still relatively small startups who are desperate for revenue.
Mark-ups or hidden deals: Agencies who operate proprietary or 3rd party trading platforms are often able to add other layers to this practice - doing deals with ad networks and publishers where they get a chunk of guaranteed inventory at preferred rates but are able to make it look like it was bought via real time bidding, thus managing to artificially inflate the price that clients pay. Often only a couple of people in the agency will know about this and be managing the practice, so it's really hard to detect and will never show up in an audit. This is a classic example of the specialisations amongst the 40 thieves - a couple of experts who understand the technology and the ecosystem and know how to use it to generate significant margins.
Non-target impressions : 10-15% of budget : Between misreported IP addresses (rampant in the China telecommunications industry where ripped off data bandwidth from one province is sold in another), VPNs and deliberate misreporting by publishers, ad networks and DSPs, ads aimed at young Chinese women in Beijing could be getting delivered to a middle aged Indian male tourist in Kunming (actually happened to me several times). Lack of 3rd party research and monitoring contributes to the possibility of this kind of malpractice.
Fake views : 20-50% of views being misrepresented by accident or design. We've all been following the story on Facebook's miscalculations, which appear to be driven more by incompetence than by intent. However, the lesson is that viewing data can be easily manipulated. Again, this possibility is exacerbated by the lack of 3rd party monitoring.
Non-performing impressions : If you've given your agency a target - downloads, views, whatever - and they achieve it on a portion of the budget by buying good quality inventory and optimising it well, they're not going to come back to you and give you the savings. To be fair, you probably haven't incentivised them to do so. They'll ensure the rest of the budget gets spent on junk inventory and impressions just so they can earn their commissions, rebates and so forth on it. Particularly when a client is not very experienced in digital media and doesn't know what a sensible cost per KPI should be, there is room for a lot of this exploitation. Because they've now demonstrated that it does, in fact, take XX$ to deliver a download, that becomes your norm and you'll continue to pay that price until someone demonstrates how inflated it is.
If you look at those percentages it becomes pretty obvious that you could be wasting almost all of your digital budget. What are some easy ways to prevent this drain?
First of all, the one thing that really helps is starting a campaign with a hard KPI. Not views or impressions but a consumer action. The moment you do this there is a theme of accountability and outcome orientation to the campaign that forces all the players to pay more attention to actually delivering something that is externally measurable. You might say "we can't sell toothpaste (or hamburgers, or cars, or whatever...) online" but that doesn't mean you can't come up with a sensible KPI that's more in your control to measure than anyone else.
Once you've done that, taking a partnership approach is better than saying "Right, bid the lowest price for that KPI and then go deliver it". Allowing your partners to work through the stages of optimisation, sharing the data and learning from it together, and offering to share any savings they generate will go a long way to getting everyone's motivations aligned.
Something that eliminates a lot of wastage is cutting out as many layers of middlemen as possible. Starting with your 'agency' - especially if it's an old school media or 'full service' agency who are going to farm your campaign out to a bunch of other people. Even a lot of the so called digital specialists don't actually optimise and run your campaign themselves - so if you can eliminate all these layers and go directly to a performance specialist, DSP or even the media owners you've clawed back a lot of the money that's just disappearing in rebate deals. You can then treat these partners as real partners and try to align everyone's interests as I've just described above. Many smaller DSPs offer a "managed service" model with a lot of transparency and over time this can transition into a pure SAAS model where you have 1-2 people managing a platform and doing campaign optimisation directly.
It's really important not just to give your partners a one dimensional KPI (like getting 50,000 downloads at $1.25 each) but also adding in some measure of quality or response rate. If you don't do this, they might deliver the target downloads by buying up lots of low price, very low response inventory so you're going down to click through rates of 0.01% which means 1 person in 10,000 clicks on your ad. That means you're annoying 9,999 consumers for every one who responds. Unfortunately CTRs don't go very high nowadays but if you put in a norm of 1% that's 100 times better than leaving it completely open, and with good creatives and better targeting it is possible to get CTRs in the 5% range for specific campaigns.
There is one other thing which might help - which is having external measurement - but this is easier said than done. So far, there haven't been a lot of very credible 3rd party measurement tools and when I've seen clients use 2 different sources we often found lots of variances between them, as well as differences in tracking between them and us where a significant proportion of legitimate delivery was not being picked up. At the minimum however, investing in some tracking and then being open to understand it along with your delivery partner would help create more transparency.
If you're spending significant money on digital media (upwards of USD 5 million a year or so) it might make sense to invest in a platform and one person to manage it so that all your partners are delivering real-time data instead of printed reports. This eliminates a lot of the opportunities for doctoring and dressing up the data.
So, in summary. Set real KPIs, find the last mile delivery partners (eliminate middlemen), align motivations, share and learn along with your partners to improve outcomes. If possible, invest in a platform where you can have real-time data and someone analysing it who can own campaign optimisation. Doing that won't eliminate all 40 thieves, but it can put a lot of them out of work...
(Note: The "Ali Baba" here has nothing to do with the famous Chinese e-commerce giant - the reference is to the story from the Arabian nights)
From November 2015 until recently, I worked in the gracious city of Taipei in an up and coming big data / mobile ad network company called Vpon. It was one of the most fun years of my life. I worked with some amazing people and learned a lot but eventually decided I needed to move back to Shanghai.
That time at Vpon, however, helped me get behind the scenes of many of the key themes in digital media today - mobile, programmatic, big data and so forth - and here's a few of the key things I learned:
1. Mobile is a great medium in some ways - but it lacks visual impact and needs to be used correctly.
Specifically, in-app mobile advertising has a degree of addressability that is unique in media - the possibilities of app detection, LBS and unique device IDs lead to an ability to target audiences based on segments as well as scenarios, and deliver the creative that is most likely to have impact at each stage. However, many clients ignore all these possibilities and use mobile in completely ineffective ways, wasting a lot of money.
2. Big data and programmatic buying offer amazing opportunities to deliver targeted messages - the right copy to the right person in the right context. Right now they're being used to drive margins for agencies and media companies.
Unfortunately a lot of the technology in programmatic is being misused as a way to buy cheap impressions in a RTB marketplace, drown consumers with lots of impressions at low CTRs and drive conversion through this barrage of cheap impressions instead of targeted, effective advertising. This behaviour is often driven by agencies who want to buy impressions cheap and then make money either by marking up or by managing a margin on a KPI like views, acquisitions or downloads - tons of cheap inventory with low conversion can still lead to a low CPA and a good margin, but that's what is annoying consumers and leading to the rise of ad-blocking.
3. Most clients are not clear what they're trying to achieve with digital - they don't know if the goal is to drive awareness / brand preference or to enrol new users and downloads.
I've seen more vague briefs in the past one year than I saw in my many years in agencies. Most of those briefs, by the way, came from media agencies. The clients who are clear what they want to do usually work directly with a DSP or digital specialist and stay away from the non-specialist agencies.
4. Most of the people who get the nitty gritty of mobile and big data don't really understanding marketing, and vice versa.
This is particularly true of ‘data scientists’ who are mostly highly academic nerds playing with their analytical toys, with no concept of which variables on their dashboards are of any use in the real world of marketing.
The few people who understand strategic marketing / advertising AND get the details of what data gets generated in digital media and how to use it are usually working at the big e-commerce firms. Most agency 'digital specialists' are basically suits who talk a good game but don't know enough about how it works.
5. Digital has become the final bastion of non-transparent practices that take away a significant part of the client budget which should have been spent on communicating with consumers.
Clients going direct can drive significant savings and ensure a much larger part of the budget gets spent on what it was intended for. Just to touch lightly on a very deep subject, there are more ways to siphon off money and more ways to deliver fake output within digital and mobile than anyone can imagine - and lots of very bright people spend lots of time on perfecting the existing ones and coming up with new ones. Some of these are very technology driven, and no audit of invoices and delivery will uncover them because no auditor has a granular enough understanding of how digital media really works. However, the ray of hope in there is that the really knowledgeable e-commerce clients usually manage to figure all this out and eliminate all or most of this wastage, so it is possible.
So what positive lessons can I conclude from all of that?
There are broadly two:
The first is to remember that 'digital' is not just an advertising medium - it's an entire ecosystem with media, marketplaces, service delivery systems and even consumer identities and personas in it. To use display or mobile as a medium with exactly the same objectives as television (reach building, for example) is to waste an amazing opportunity. Even if you're a brand that sells entirely offline and doesn't have an e-commerce model, think about what you can do better or different using the digital ecosystem - don't just measure reach or GRPs on it.
The second is to encourage clients who spend significant money on digital to hire an expert in-house, preferably from an e-commerce firm or a performance agency campaign management background. Bypass the 'traditional' agencies and work directly with the plethora of specialist DSPs /performance agencies. The key benefit of this will be reducing the margin that otherwise gets diverted to agencies, but apart from that an in-house expert dedicated to one client will always do a better job than someone at an agency juggling multiple clients.
To end on a somewhat practical note I thought I'd list the 3 worst uses of mobile in-app advertising that I've seen in the past year:
1. Sending consumers to a 'brand' website.
Remember, a mobile screen is usually pretty small (most AD networks can tell you what proportion of their audience is on a phone / tablet etc. not to mention splits by OS, so you can figure out what proportion are operating off a small screen) - sending them to a brand website, even optimised for mobile, isn't that great a value. Better to do this on PC - although this only makes sense in the first place if your website adds some value to the consumers life. Most I've seen fail miserably on this count.
2. Running a 30 sec TVC on a mobile.
Again, it's a small screen. Often when people are using an app and rapt in it you're just interrupting them, and if they're not on wi-fi, using their data subscription to send them a video ad they didn't ask for and, in any case, can't see properly on a small screen. Given the degree of segmentation and scenario targeting you can do on mobile, why not serve something more specific instead of using the same 30 sec mass reach copy?
3. Trying to generate leads via in-app mobile ads that lead you to a website - and then making the process overly complex. Just get a name and an email ID and leave the rest for later! I've seen some landing pages with over 20 questions on them - who's going to fill all that out, especially on a small screen?
Don't use digital or mobile media just because they seem to have become de rigeur for any modern marketer. Use some common sense, figure out what you really need to achieve and how to use the appropriate ecosystem to achieve it.
Glossary of acronyms:
(I just had my first guru in advertising -Sumit Roy- point out that I used a lot of acronyms that people may not understand, so I'm adding a short list of acronyms and what they mean. Thanks for the catch Sumit.)
LBS: Location based services - in this context it referred to the ability of an Ad Network to track the location of mobile phone users and build contextual targeting models around their location information
CTR: Click through rate - the ratio of people served a (clickable) ad who actually click on it and follow it to a link (I didn't think I needed to include this in the glossary but Sumit assured me I did!)
RTB: Real time bidding
KPI: Key Performance Indicator, or in simple terms, a specific metric against which a campaign is to be measured for effectiveness
CPA: Cost per acquisition - (usually in an e-commerce context) the cost to acquire a new customer
(First published on Linked In in January 2017)
Let me say this right up front - I don't think it's right for agencies to make money in ways that they try to hide from their clients. I don't condone that for even one micro-second.
However, if clients read that line and think "Yes, that's our money and we should get it all back" that's wrong too. There are fundamental reasons why a lot of these shady practices started, and to stop them you have to fix both sides of the equation.
Agencies are not unscrupulous villains exploiting innocent clients and stealing money from them - clients can be pretty villainous too.
As someone who spent 20 years in agencies and is now heavily involved in the programmatic buying industry, I wanted to try and offer a balanced perspective.
Many clients today ask agencies to guarantee costs. If the agency can't meet the guaranteed cost, it usually has to foot the bill for the difference. I've seen this happen before, on a client I won't name, the agency gave up their fee and paid a penalty of around 1.5 million USD in addition.
This is not at a 'total cost' level. Costs are broken down into great detail and the expectation is that the agency will meet EVERY number. If they go above, that's their problem, but if they go below, the client expects to make the saving.
That's fundamentally unfair and imbalanced. The reality is that when you're negotiating a large number of deals, some will come in above expectations, some will come in below. A client who's willing to treat the cost estimates as estimates and roll with the punches can expect transparency. A client who wants all the upside but none of the downside deserves a swift kick in the pants.
Some clients understand the realities of media deals. They set targets in collaboration with the agency and offer support and resources to help them in negotiations. P&G media executives would participate in media negotiations if we asked them to, they would adopt the positions we sometimes advised them to in meetings, and when we occasionally blacklisted a channel or media vendor in order to play hardball with them, the P&G media guys backed us up with their colleagues in brand management. Almost every single negotiation I remember from that time ended up with gains that we did not expect, far better than any target we may have set ourselves.
In those situations, there may be a goal to reach, a target cost. However, success or failure is a shared outcome between the client and the agency. The agency gets paid a fair fee for making it's best effort, so it's interests as a business are protected.
This kind of client has the absolute right to know every detail of every media cost, and they usually do right from the start, because they were part of the process of signing the media deals.
The problem with this whole process today rests on three things:
1. People from purchasing who don't understand the dynamic nature of media pricing (yes they understand pricing but do they understand how ratings / impressions fluctuate? A widget is always a widget except when it's a media widget) and try to enforce commodity pricing models on this industry
2. Media auditors who get paid for poking holes in what an agency has done - and that's always an easy task, isn't it? The money that gets spent on fees for these guys would be far better used as an incentive payment for an agency.
3. The unwillingness of agency management to say no to any business or to have difficult conversations with their clients. This is driven by the unnecessary public listing of agencies leading to unrealistic financial goals and pressures, and it's not helped by the fact that most senior people in agencies now are paper pushing bureaucrats who wouldn't recognise a consumer insight or a creative idea if they ever saw one. More on that in an earlier post here.
While on that topic, let me make a somewhat unrelated point.
Marketing is not a core business - it's an ant in relation to the rest of human endeavour.
Advertising is the bum of the ant - the piece that comes in the rear after the vanguard decisions about product, pricing and distribution are made.
Media is still (unfortunately) a dimple on the bum of the ant - the bit that gets decided (usually) last.
Media auditors are an addendum to the media agency business - scavengers living off a tiny percentage of that revenue model - so in my book, they're a pimple on a dimple on the bum of an ant.
Let's stop losing ourselves in the intricacies of this business and remember a few things:
1. A saving of 10% of ad-spend is usually less than 1% of a company's overall cost - there are far better areas to focus on cost reduction which are more impactful for every hour of purchasing effort spent.
2. Rather than commoditising media and expecting CPT or CPM guarantees it would be better for clients to focus on outcomes. Work with agencies to jointly negotiate deals that are not just looking at cost but also evaluate the effectiveness of the media buy.
3. Recognise that agencies need to make some money too - when they make an honest effort for a client, they deserve to make back their costs and a reasonable margin, even if the outcome wasn't as good as expected. If you chose an agency badly, that's not just the agency's fault.
Making media auditors the policemen is a very bad idea - because they're not neutral. A media auditor who says "Your agency is doing a bang-up job, don't change a thing" will soon be out of business. A media auditor who points out problems, areas for improvement and offers more consulting hours to diagnose the problem does much better. A media auditor who rings alarm bells and calls a pitch does best of all.
So, in summary:
1. You want transparency? Fine, then don't ask for price guarantees (conversely, you want a guaranteed price? Then don't ask for transparency and plan on changing your media agency every year). Participate in the media buying process, be invested in the outcome, be a partner, not a pain in the ass.
2. Stop wasting money on media auditors - use the money to reward your poor, hard-working agency for their efforts.
3. Stop acting surprised when people produce voluminous reports to say "agencies are keeping money from their clients". Of course they are, did you not know that already? Of course it's bad, but you contributed to the pressures that made that happen. Now think about what you can do to make it better.
Of course, agencies are far from blameless in this whole thing. Pitching business at rates they can't really attain, undercutting each other and enabling the greed of client purchasing departments... you made your own bed boys and girls, you contributed to this situation too. It's not that hard to change it - it starts with trying to have a real relationship with your client - tell them about your business and what it would take to be truly transparent with them. Say no to pitches which are all about filling up numbers in an excel spreadsheet (giving enemas to very small flies, in my book).
That's what it boils down to, ultimately - creating a real partnership. Partners don't exploit each other, threaten each other with breaking the partnership for small problems or quietly do things behind their partner's backs, hoping not to get caught.
(First published on Linked In in June 2015)