Case Studies

 

Case Study One

'Make It About Data, Not Systems'

 

Author: Mark Darlow

 

Challenge:

Any organization once it has installed a key system to manage a set of functions tends to build other systems around these as their business changes and adapts. It seems much easier and less disturbance to the business to add on small applications and databases to a main system. The issue is over time, these starts to create convoluted workflows and multiple data entry points for common data. Costs of operation thus creep up.

This paper looks at how one captures this process, its implications and reviews how to ensure one continuously optimizes all applications within an ecosystem to deliver the most efficient workflow for the business.

 

Discussion

In the discussion that follows we look at how deployment of more and more systems impact the cycle of re-entry of common data caused by deployment of ever more applications to service the business versus an approach that looks at data master-sets that are not re-entered, but data that is shared.

 

Lean Manufacturing

The concept of ‘Lean Manufacturing’ is now very widespread, the idea that all tasks and points in the manufacturing process have a cost to them and as tasks and functional transactional points multiply, thus the cost rises.

Below is a diagrammatic of how ‘Lean’ works in manufacturing. Lean manufacturing or 

lean production, often simply ‘lean’, is a systematic method for the elimination of waste within a manufacturing system. Lean also takes into account waste created through overburden and waste created through unevenness in workloads (or workflow) as we might think of it. Toyota developed a second approach to lean manufacturing, in which the focus is upon improving the "flow" or smoothness of work, thereby steadily eliminating unevenness through the system and in doing so ensuring 'waste’ reduction.

Source Wikipedia

 

System Ownership – Not Data Ownership

Most people are familiar with the problem. Across an organization, teams and departments have responsibility for functions. Allied to those functions will be datasets, program data and material data are two examples. It tends to be the department which makes the claim for their own systems, either through direct ownership of a budget of sponsored by an overall budget from IT. In the data examples given, there is commonality to some extent, but the functions are very different. Program data for example, contains information about the program event, who is in it, what the story is, perhaps certification information. All this data is used by scheduling staff, those responsible for creating EPG data sets and so forth. At the material side, formats become important, state of quality assurance, location of the media, readiness for transmission, timing information. This is different information, managed by a different group with different priorities, but their media file needs to be linked to a program and episode, so at this point two systems now have common data of program name.

 

 

Multiply this out by the number of event types and business functions that are associated to a program and one can see system and data entry becomes exponential. Examples can be sales systems, contract management, financial systems, promotion management systems, resource systems, non-linear scheduling. The number often depends on the core systems in place. The media industry has an opportunity to learn from manufacturing industry in how we manage data and workflows.

The diagram below shows how this manifests itself across media organizations.

 

 

 

 

 

 

With each department using its own system, and entering common data multiple times, often with the same data created different ways through spelling error or style of entering data, staff working closer and closer to the transmission point have to then assimilate and correct much of the data.

Hence the organization is faced with workflows such as that below, where small amounts of data are added upfront and then as the content and media all progress through their dedicated functions and workflows at different times pursuant to their workflow cycle, the same data is multiplied throughout the system. In the example, 10 business functions all potentially using different systems, whether it be a database driven system or excel sheets require data such as program name. Linking this data together becomes complex, is most often manual in manner and becomes a very uneven workflow with cost building throughout the process associated with data entry.

 

Data Ownership – Not System Ownership

If we reverse the system and data dynamic and make the data the item to own, we start to leverage data entry performed upstream. Even if the data has to go to another system, automating that through a messaging system or ESB (Enterprise Service Bus) still delivers the same saving in time and data accuracy.

The business generally starts with creation of contracts and content metadata. These items are known and managed well in advance of any operational work on scheduling or media management, so leveraging the effort of this group into the system(s) used by others makes for the start of the “lean” approach to data management.

The result starts to look something like the diagram below.

 

The organization as a whole does need to work out where responsibility for data sits and come to terms with the concept that one department own the data that in many cases another department requires. It also means that good communication needs to be in place between departments to cover inevitable contingencies where data is missing or is not as expected. A good level of communication also provides a quality assurance to the data being entered as is has a high visibility to many people.

 

The resulting impact on the organizations workflow look something like the diagram below. As one can see, there are less steps in the process, the workflow has a whole is much more even. This is the key part of a ‘Lean’ approach, creating even workflows, which in term cut out unnecessary steps and deliver a lower cost of operation.

      

Conclusion

Our industry can learn much from the ‘lean’ manufacturing philosophy. In essence data represents the pieces and parts of the media production process and thus creating even data entry where the data is entered at the best possible time and shared across the business systems deployed, delivers the most even workflow, which in turn delivers a simpler, nimbler and more flexible operation at lower cost.

Managing this process though requires an organizational philosophy to embrace this approach across all departments, a willingness to share data as well as to rely on others for the data required. It is simple to conceive, but requires a collaborative culture to implement.

Wikipedia Link: https://en.m.wikipedia.org/wiki/Lean_manufacturing

Case Study Two

Development Proposal

Blockchain For Media Contracts

Challenge

The existing business process between seller and buyer of media and the terms delivered via contract is reliant either on a matter of trust or at some point an invasive audit to confirm compliance with the contractual terms.

For the buyer, the management of a contract might include three or four systems and increasingly and outsourced supplier for both playout and non-linear content delivery. There is a strong argument that Blockchain technology, increasingly that referred to as Blockchain 2 and the smart contract capability could provide the technology solution to this problem.

 

Discussion....

Negotiation

The process of negotiation is at present driven by paper rather than any system. Whilst systems exist such as “Media Maestro” to manage the details of a contract for a buyer or even a seller, there is no standard for information transfer nor at present any automated transfer. All contracts are paper form, delivery as pdf from seller to buyer etc. with terms agreed and signed through this mechanism. The data is different, but the process seems no different buying a program that it does buying a house.

 

The Buyer’s System Infrastructure and Processes

Most buyers’ systems will include three or four dedicated enterprise level tools and in most cases a myriad of spreadsheets and smaller databases often built by individual users. A number of larger organizations have invested in systems that manage the legal terms of contracts, territories that the contract covers as an example. These systems though tend to be isolated and are not part of the content delivery chain.

The chain starts with scheduling systems which incorporate the contract terms, as per the workflow below. Systems such as Sintec, MediGenix and Imagine Vision deliver the functionality described. Data in entered into these systems covering the cost of the contract, payment terms, terms of use, internal accounting rules all of which are updated as the content is used. Unused content delivers a “stock” value and used content is amortized and both update the accounting systems ledger.

The scheduling process is linked to contract utilization and as the playlist is generated and returned to the scheduling system as an as-run schedule, the events are marked for update. Typically, an end of month accounting process than updates all events within a given time frame and sends update stock and amortization positions to the financial system.

Within the financial system, the update monthly contracts management process tends to trigger payments to the seller, which are either date driven, or content usage driven.

In all cases, the seller is reliant upon the buyer to deliver payment and confirm agreed contractual usage of the content on trust as they have no visibility into the buyers’ systems and no automated notifications.

The Workflow Process Of Contract Management By A Media Operator

 

Blockchain And The Opportunity

Smart contracts can be considered contracts where the key terms are codified as business logic and operate on a blockchain or distributed ledger. They automatically verify, execute and enforce the contract based on the terms written in the code. One of the challenges but also the opportunity is to provide a UI layer that would allow a user to configure the contract from an a-la-carte menu to create unique contract terms.

Additionally, these Smart contracts can be partially or fully self-executing and self-enforcing. Smart contracts can be used to exchange anything of value. In this way, individual plugins could be created to connect different systems and system types into the contract logic and ensure that the systems in the legal, scheduling and financial chain are updated with information that is both recorded on the blockchain, timestamped and can’t be altered. This creates an ideal platform for contracts to update based on scheduling and financial updates.

The workflow identified above outlines a use case. In this case the contract negotiation can be captured and versioned until the point there is agreement to the terms. This then becomes “active” or approved.

The second part of the workflow is data exchange to the scheduling system of the buyer. The key data that is standard, timeframe, price, currency purchased in, regions and distribution type can all be passed via a messaging plugin that may have to map contracted terms data to the terms used by the scheduling system. Verification that the data was sent and received should be held on the contract and Blockchain.

Once the program event is schedule and transmitted, just as the as-run process updates the scheduling system to mark the events available for month-end processing than the contract itself can b updated with the event transmissions. A plugin would pick up the key data from the scheduling system and place that within the contract on the blockchain.

This data is visible on the blockchain by the interested parties and this provides transparency to the effect that the scheduling processes have honored the terms of the contract. A plugin could also update the system of the seller if they used internal systems to record contract usage.

Following on from this, payments can be prompted from the Blockchain and update the buyer’s financial system to trigger the actual payment details recorded in the contract so both buyer and seller can see the financial transactions related to the contract.

This process is not unique to the media industry, many of the industries utilizing blockchain technology will be using smart contracts. When a smart contract is run on the blockchain it operates automatically. If the conditions of a contract are met, payments or value is exchanged based on the terms of the contract. For media, this will tend to be monetary and runs / usage value. Likewise, if conditions in the contract are not met, payments may be withheld if written into the smart contract.

Smart contracts run as they are programmed on a decentralized network of computers on the blockchain removing risks around unauthorized changes, fraud, server failure or non-compliance of the terms of the contract. The contracts execute automatically, exchanging value and payments between people without the need for external audits to verify or enforce terms.

There are a number of companies that are using smart contracts in various ways, Ascribe, UProov and BitProof three examples of tracking and verifying changes to contracts.

Indeed, Ascribe, is a start-up in the art industry has a workflow that could find its way to supporting all media production. Issuing numbered artwork in its digital form uses a blockchain to trace back all original creations and transactions within those creations. This would be useful for initial creations and the versioning of content that is widespread across the media industry. It provides a full DNA trace of the content and this should be the aim of this system, allowing the sellers to comprehend exactly what the buyer is doing with their content once it is shipped to them           

 

Conclusion

When discussing the future of the Blockchain, the term “Blockchain 2.0” is commonly used to describe the next step in the evolution of blockchain technology. Decentralized apps and plugins along with smart contracts can provide much needed improvements in the efficiency of contract management both internally for the buyer but also between buyer and seller as contracts are versioned and stored. The transparency the system provides to both parties is a much-needed development and as media delivery prolificates across more distribution mechanisms will provide an automated proof of compliance that will otherwise be near impossible to achieve.

Amongst those familiar with the Blockchain technology, they consider that smart contract capabilities take use of blockchain technology to exciting new levels. The future of the blockchain will revolve around smart contracts and dApps.

References

Blockchain: Ultimate guide to understanding blockchain, bitcoin, cryptocurrencies, smart contracts and the future of money." by Mark Gates

Case Study Three

The Business Survey

What Is A Business Survey?

A business survey is traditionally the process where a client and vendor engage in a discussion that allows the vendor to determine how the new system will be deployed and the behavior of it, which functions to switch on and off. For any sales system this will entail a lot of questions about, the trading methodology to be supported by the system and recommendations of system settings to deliver against the trading “policy”.

Any vendor is focused on questions and advice that enables them to serve their client best and in this case, it means asking the questions that map trading policy to system settings. Examples would be if spot booking were the sales methodology, would the client want to use minimum price on the break function or just recommended? This would be a switch on the system to enforce a function to only allow booking that meet the minimum price.

You can see from this example, that the term business survey relates to how the system will support the business.

 

The Crucial Question

The Sales System is a complex piece of software, arguably unique in how is services the advertising sales business. One of its great strengths should be is its ability to service multiple trading models. Spot booking, spot booking with ratings, trading against a CPM, trading against a discounted CPM (which allows for more optimization of the inventory) and now of course trading against audiences across multiple delivery platforms. So, the key question becomes, not the Vendor question of “how do you trade” but much more a business question. What are your strategic objectives and what is the best trading policy to meet that? To ask that question you need to understand capabilities of the system and the impact of the various methods on trading practices of the organization. MediaBond focus on these questions with the client, so that when the vendor conducts their survey, the client is increasingly aware of what it will mean for their sales policy implementation, possible market communication that needs to take place and impact on the sales team.

 

Operational Efficiency

The Sales System should offer many functions that provide operational efficiency to a booking team and sales organization. Using the example above, a predicted rating applied to a break may determine the break price and manage the booking entry, outlining the predicted CPM. Subsequently, after transmission the rating changes to reflect what really happened, the CPM changes to show actual performance and across the break you can look at yield, dollar value the break delivered, CPM by audience and so forth. The performance of the yield however is subject to how good the rating prediction was, which in turn is subject to volatility of the rating service. Whilst the booking process and management of it is extremely efficient, the yield is more a reflection of luck than planning. The Achilles heals of this booking process is well understood be statisticians, predicting the specifics of one event is much harder than predicting a trend. Macro-economics versus Micro-economics. Sales systems should offer both.

 

The Use Of “Big Data”

The term used these days for an ability to draw conclusions from a mass of data.  Some Sales Systems achieved this for many years as part of its approach so setting sales policy.

It represents a different way of trading however, by using data from previous months’ ratings, one can look at overall delivery against the way deals and campaign boking have stacked up. This is management of supply vs demand and how efficiently you are likely to be able to meet the demand. Some examples will suffice.

Do the ratings of booked spots match the natural day part delivery? For example, Adults 16-34 have a natural delivery of 30% daytime, 25% early peak, 35% late peak, 10% late night. Bookings how that ratings taken by % is 20%, 20%, 45%, 15%, which tells us the organization will run out of peak airtime to sell and be left with open time in daytime. As time is used up, the yield of the individual break of the audience required vs a base audience declines and will require a price supplement to reflect this. As all advertising agencies work on a budget, CPM and thus fixed delivery of ratings, spot pricing becomes irrelevant as a measure, agreed CPM, agreed ratings and the specific budget are the contract, break prices just the booking methodology.

If at the outset, the annual deals done do not reflect the expected demographic delivery of performant ratings, then the supply-demand equation will the wrong from the start. The organization can use this data to determine is sales trading approach, selling available audiences at a discount and those it will find hard to deliver demand a premium.

For example,

Monthly delivery of All Homes, 1,000,000 ratings

Monthly delivery of Adults 16-35 600,000

Daytime daypart delivery of All Homes 330,000

Daytime daypart delivery of Adults 16-35 220,000

Break delivery of All Homes, 10.2 ratings

Break delivery of Adults 16-35 7.8 ratings

This leaves us with the following performance metrics.

	Homes	Adults              16-34	Conversion	Performance Index All	Performance Index Day
All Time	1,000,000	600,000	0.60	-	-
Daytime	330,000	220,000	0.67	111.11	-
Break	10.20	7.80	0.76	127.45	114.71

 

 

From this we can determine that the daypart delivers better than average, and that the break itself delivers better that both the all-time average and daypart average. This kind of data would be used by the sales system to determine what spots go into the break, and sets a level of minimum performance criteria. Only spots of 115 Index go into the break.

The result is more efficient slotting of airtime, usually more peak time is made available for sale. The difference between the two models if the former leaves agencies in charge of selection over the inventory, whilst that latter gives the media operator more control, whilst still meeting the contractual obligation of the deal. Deciding which model to pursue though means very business and organization specific questions rather than questions relating to anticipated system behavior.

 

The Business Questions

What are the expected market trading practices?

Does the client want to confirm with the standard or do they have an appetite for innovation?

Do they overtrade in peak time and is this limiting their revenue growth, i.e. too high price vs supply available?

How is their sales organization setup, do traffic do booking on behalf of sales?

If so, is there an appetite to change the organization, sales assistants do bookings supporting sales executives?

How are deals struck, annually and are they against audiences and fixed prices or discounts against a station price?

How does the market and organization want to manage pre-emption, spots cancelled, go into a wait list etc.? Do agencies need to know of pre-emptions or as long as they are replaced in similar dayparts and campaign parameters met all is OK?

Do agencies and client want to use budget only at campaign level or spot pricing. The benefits of the former can only be delivered based on trading agreement, but reduction is administration is considerable for both parties, e.g. no spot matching required.

 

The Sales Trading Strategy

The questions above determine the sales trading strategy and that is what the vendor will be asking about such thy they can ask relevant questions concerning system settings and its expected operational behavior.