Advertising Technology

Emerging technologies always encourage scrutiny and critical analysis, and advertising technology (ad tech) is no different. This discipline has been around for a few years, but it's only recently caught the attention of savvy ad tech vendors. In the era of big data, we've recognized having ad tech company relationships makes us more powerful and attractive to clients. The age of social media has forced agencies to investigate innovative ways to interact with relevant users, rather than relying on typical broadcast or digital media buys. The first step to leveraging ad tech is understanding that it is the umbrella term for the software and tools that help agencies and brands target, deliver, and analyze their digital advertising efforts. If you've ever scratched your head at the terms "programmatic" or "omnichannel," you've likely already heard a little about what ad tech does (though you may not have even realized it). Programmatic advertising, for instance, buys target audiences instead of time slots: Think about buying ad space that reaches a particular demographic wherever it is instead of buying a prime time TV spot and hoping the right people are watching. Omnichannel marketing reaches target consumers across all channels -- mobile, video, desktop, and more -- within the context of how they've interacted with a brand (those first seeing an ad will receive a different message from those who have engaged with that brand a number of times). Omnichannel and programmatic aren’t the only two tools within ad tech, but they are two of the most revolutionary.

If you’re plugged into ad tech, you’ll surely know that one of the topics around the industry is supply path optimization (SPO) which is the practice of pruning bid requests through algorithms to make smarter buying decisions and reduce infrastructure load and the next evolution of programmatic. These algorithms often look at the supply-side platform (SSP) publishers work with to analyze traffic patterns and win rates to assess which publishers provide the most direct, toll-free path to their inventory. One way that publishers are engaging with the SPO movement is through ads.txt (Authorized Digital Sellers) whch is an initiative from IAB Technology Laboratory. It specifies a text file that companies can host on their web servers, listing the other companies authorized to sell their products or services. This is designed to allow online buyers to check the validity of the sellers from whom they buy, for the purposes of internet fraud prevention. 

More advertisers think log-level data is a remedy for many of programmatic’s woes. Originally, we used the data to check that nothing shady was happening to our money, whereas now we’re using it to funnel our money through cheaper, better-performing ad tech vendors. But getting this data isn’t easy. Log level data is granular data that is collected about every single ad request. It can include everything from the designated market area (DMA) associated with the ad request to the specific reason why a bid was blocked. Most data publishers and advertisers can access within an advertising platform is aggregated. On the other hand, log level data is completely raw and can be used to create advanced reports not typically available within a platform. Log level data allows publishers to optimize their inventory’s performance from an ad call to an impression. Publishers can evaluate this data for a full perspective regarding how buyers are bidding on their inventory over time. If a publisher sees pockets of peak demand when an increase in bidding occurs, they can adjust their inventory prices accordingly. With a complete picture of their demand, publishers can optimize price floor and price rules to balance price against demand, ultimately leading to additional revenue and a higher fill rate. Publishers can also use log level data to optimize their traffic acquisition strategies. For example, a publisher may need to decide if they should attempt to drive more traffic to a desktop site or in-app content. If the publisher knows which inventory has higher demand at particular days and times, they can drive traffic to the best performing content.

Large brands and agencies have invested in data intelligence teams to help shed light on what has been a traditionally murky programmatic supply chain. Advertisers can review auction-level data to understand the full scope of the supply and demand “ad tech tax” they are paying for a single impression. Evaluating this data also allows advertisers and brands to whittle down their supply partners and develop closer relationships with a limited number of SSPs and exchanges. Stronger relationships with fewer partners can help improve brand safety and inventory quality. One of the most important things to understand about log level data is it is not delivered as a report. Any publisher or advertiser hoping to gain additional insights from log level data will need their own business intelligence software to gain any insights from these files. As the industry pushes towards full transparency in the programmatic supply chain, log level data will continue to be an essential piece of the puzzle for both publishers and advertisers.

Content Management

At first glance, “product content management” seems like another meaningless buzzword. There are a lot of vendors touting their product content management systems. There are very few of them actually talking about the practice of product content management. If vendors only use the term in an attempt to differentiate from the product information management marketplace, but still basically sell PIM, then what does “product content management” even mean? Well, fear not! If you have a business that maintains a wide range of products across a variety of channels, it’s important that you deliver consistent product information. However, it can be challenging to manage so many different products. You may find that your branding, marketing, and sales efforts are inconsistent. Know that you are not alone, many encounter the same issues. Product content management is actually a merchandising discipline that is worth understanding. It incorporates aspects of retail merchandising, digital marketing, and content strategy into one discipline. Product content management is commonly defined as: “The tasks of content creation, aggregation, categorization, scheduling, staging, publication, and syndication. Beyond the core product content assets, it also incorporates the assignment of attributes such as category, price, and promotion eligibility.” Before we get into the meat of it, let’s talk about two important marketing disciplines: content strategy and content management. These terms may sound analogous, but they are not. They are two distinct, but related, marketing disciplines.

At its best, a content strategy defines: key themes and messages, recommended topics, content purpose (i.e., how content will bridge the space between audience needs and business requirements), content gap analysis, metadata frameworks and related content attributes, search engine optimization (SEO), and implications of strategic recommendations on content creation, publication, and governance. Content strategy is the driving strategy for any successful website. It’s the discipline of deciding how you are going to use your web content to achieve your organizational goals. Content management is the discipline of facilitating the content lifecycle–you know, that content that was derived from your content strategy. How do you create it? How do you manage and/or maintain it? How do you publish it? Are there approval workflows in place? Who is responsible for what? Content strategy is the–well–strategy. Content management is the execution. You’ll often hear the term “content management system” to describe the technology platform used to facilitate content management. These come in all shapes and sizes from the free and hugely popular WordPress all the way to six-figure enterprise systems like Adobe Experience Manager. Remember though…content management is the technology and process. A content management system won’t just magically make it all work. To execute a successful online strategy, some degree of content strategy and management are necessary. Smaller companies may be somewhat informal about these disciplines, but you can be assured they are or should be thinking about it. Product content management (PCM) simplifies many of the tedious and time-consuming product-related tasks. You are able to quickly update and manage all your products, which is essential with a large or growing product catalogue. Here are some of the ways that a PCM allows you to do that:

- Control all your product content across all sales channels from one place

- Consolidate all product content and information and provides accessibility and manageability for your users

- Categorize and classify products, manage your catalog, and even import supplier catalogs

- Easily add new products and specials like bundled offers online

- Give product content a consistent look across all your sales channels

- Provide shoppers with a compelling customer experience that is consistent and engaging

While generally accepted among us as digital marketers, content strategy and content management are a little tricky to apply to a retailer’s specific business model. The ideas are still valid, but they’re a bit too abstract when taken at face value. This is especially true of content management. To a retailer “content strategy” almost becomes retail merchandising. It’s the discipline of determining how you want to bring products to your market. It’s just missing a few technical details, like search engine optimization (SEO) strategy. But, content management for retailers is a different story. When you read the literature on content management, it isn’t written to the retailer specifically. Retailers are also challenged with managing a product assortment and vast amounts of product data to describe that assortment. This is something that other business models do not face. There is no such thing as “merchandising” in any business model other than retail. Retailers must constantly source new products that will satisfy their customers. They must combine their separate supplier catalogues into a master catalogue. They must enrich that catalogue with descriptions and attributes that explain the products, while maintaining the desired brand experience. To a retailer, the products are the content. So, the retailer has to fight two demons: the challenge of maintaining proper product data and the challenge of executing a well-managed content. This is no easy task.....Content management systems, like WordPress, are therefore not built for retailers. They are not built to facilitate the processes of a retail business. They are built, mostly for managing content-driven websites, like blogs and marketing sites.

It is technically possible to run an eCommerce business on a basic content management site. And, some retailers have even tried to do it. But, the technologies and the processes they facilitate are so different than what the retailer needs, it winds up costing a lot of time, money, and mistakes to customize the system. Content management is just simply not built for retailers! As a result, many retailers either spend millions building their own homegrown systems, designed to execute their processes. Others simply ignore the issue, move to an eCommerce platform like Shopify, and neglect to think in terms of “content management”. Neither of these is an effective or scalable strategy! What retailers really need is a version of content management, adapted to their specific needs. They need best practices and technologies that are specifically designed to help them execute their content strategies–their merchandising strategies. PCM is also a hybrid discipline, combining concepts from retail merchandising, product data management (sometimes called master data management), and content management. It is exactly what retailers need to execute the strategies they define. PCM incorporates data management disciplines to ensure that your master product catalogue is always accurate and always adheres to the brand.

How do you combine cataloges from multiple suppliers or manufacturers? Their data quality varies. Their data structures vary. Yet, retailers have to pull them together into a single cohesive catalogue. What are your practices for ensuring data cleanliness and auditing for accuracy? It’s too difficult and expensive to manually check every item in a huge product catalog, so retailers must define ways to perform ongoing monitoring. How do you link products, so that relationships, collections, assortments, and kits are appropriately represented by the data. This isn’t just a housekeeping exercise. This is required if you want to effectively personalize the shopping experience or make strong product recommendations. Rich metadata and digitally linked product data are not optional. PCM includes the execution of product lifecycle management workflows for bringing in new products, maintaining existing ones, publishing them to sales channels, and sunsetting products. In larger retail organizations, many people may touch the same product over the course of its life. Merchandiser A might source the product, adding it to the catalogue. Merchandiser B might define the brand-friendly product descriptions. An eCommerce team member might worry about publishing. Merchandiser C decides when the product is dead. PCM should include workflow definitions and tools to help execute them. It should ensure that products can be managed in a scalable way, helping to avoid a giant mess of data. Just like a content management system facilitates publishing web content for the world to see, product content management includes publishing products to sales channels. The publishing aspect of product content management may (or may not) happen in multiple technologies. For example, you could execute PCM inside your eCommerce platform, like Shopify, etc. You can bring in your products, update their attributes, and publish them to your webstore. But, this single-technology approach has its shortcomings....

So, what you end up seeing is a multi-tier technology architecture for PCM. One system for managing the data, one or more for publishing the data. Your exact technology architecture may vary, based on your needs. But as a discipline, PCM must include publishing that product content to the world. You may have heard of product information management (PIM) solutions and wonder how they are different from PCM solutions. And you will probably find that there is quite a bit of confusion about how they differ. As described, a PCM offers data enrichment and product information management (PIM) solution offers the same as a PCM, with some additional components like aggregation and data quality. Aggregation allows you to collect and combine product information from various sources and in various formats into a single source of information. Data quality refers to the ability to normalize and remediate any problem data. With a PIM, you will typically pay more for the added components. Despite the growing importance of product content management, many business owners are still not familiar with its uses. If you are still managing your product content manually, it’s definitely a solution to consider. A PCM can have a significant effect on your efficiency and scalability, which in turn, helps your business grow and prosper. To be honest, a lot of really big, well-known retailers just aren’t thinking about eCommerce in this way. And, they should!

Distribution Solutions

Due to technological advancements, there is increasing adoption of streaming solutions and services on a wider platform which helps in better marketing and branding of products. In 2018, streaming platforms had grown to over 200 million monthly active users across the world acoording to some research reports. The streaming market is expected to witness significant growth with the U.S and Canada anticipated to drive the growth of the world's streaming market. This is owing to the presence of large number of established players in video streaming market. In addition to this are well-established infrastructure which allows higher penetration of mobile devices which ultimately provides high speed connectivity and is expected to be a major factor for the growth of the streaming market. Majority of the companies are adopting streaming solutions and services for their marketing and branding activities which ultimately helps in driving the market growth of streaming market.

Kubernetes, originally designed by Google, and now maintained by the Cloud Native Computing Foundation, is seen as being faster and a safer option for streaming services. Kubernetes is an open-source container-orchestration system for automating application deployment, scaling, and management. It aims to provide a "platform for automating deployment, scaling, and operations of application containers across clusters of hosts". It works with a range of container tools, including Docker. Most streaming service consist of a mesh of different API services all written in node.js as well as the go programming language. Some streaming platforms are also deployed on Amazon Web Services (AWS) and built using EC2 virtual instances, set-up with auto-scaling capabilities to help handle demand using load-balancers to distribute traffic.

Many streaming services use Kubernetes-based platform or infrastructure as a service (PaaS or IaaS) on which Kubernetes is deployed as a platform-providing service. Many streaming services also aim to empower creators and enable a really immersive experience for their consumers. Streaming services adopted microservices and Docker in the early days with containerized microservices running across a fleet of Virtual Machines with container orchestration systems. As the streaming services grow, it becames clear to those vendors that had small teams working on these features that their technical operation were just not as efficient as those who adopted platforms that were supported by bigger communities. This sparked the growth around Kubernetes which was been constantly developed with more features that added velocity and reduced cost, as well well provided the industry with best practices and tools. At the same time, the team wanted to contribute its expertise and influence in the flourishing Kubernetes community. Those services which migrate to Kubernetes find that it fits very nicely as a complement to their existing platform as well as a replacement. Among the reasons why they chose to go with Kubernetes is the improved utilization and introspection capabilities that the technology provides.

Some of the biggest streaming services running on Kubernetes take about 10 million requests per second as an aggregate service and benefit greatly from autoscaling. Before, teams would have to wait for an hour to create a new service and get an operational host to run it in production, but with Kubernetes, they can do that on the order of seconds and minutes. In addition, they reported that with Kubernetes’s bin-packing and multi-tenancy capabilities, CPU utilization have improved on average two- to threefold. Some services also found that many problems with their services were not caused by Kubernetes but were there all along and Kubernetes made them more visible. The streaming market is rapidly over 17% of Compound Annual Growth Rate and expected to reach at approx. USD 82 billion by the end of 2023. The prominent players in the streaming market are - Netflix (U.S.), Adobe Systems Incorporated (U.S.), Ustream (U.S.), Amazon Web Service, Inc. (U.S.), Akamai Technologies (U.S.), Microsoft Corporation (U.S.), Apple , Inc. (U.S.), Google (U.S.), Hulu (U.S.), and Cisco Systems, Inc. (U.S.) among others......Segments of the streaming market are based as the following: By Streaming Type - Live Video Streaming and Non Linear Video Streaming; By Platform - Laptops/Desktops, Tablets/Smartphones, Smart TVs, and Gaming Consoles; By Deployment - Cloud and On-Premise; By Solution -Pay TV, Internet Protocol Television, and Over-The-Top (OTT); By Service - Training, Support, Consulting, and Managed Services; By Revenue Model - Subscription, Rental, Advertisement, and Retail; Lastly By Vertical - Healthcare, Education, Media & Entertainment, IT & Telecommunication, Retail, and Government;

North America is estimated to account for the largest share of the market, whereas Asia-Pacific is projected to grow at the fastest rate during the forecast period. The major growth in video streaming market in North America attributes to the technical advancements and increasing use of mobiles and tablets in that region. These are “affordable devices capable of supporting streaming services, and affordable, ideally uncapped, internet data. These are seen in waves across the world were developed regions are amongst the first to get ubiquitous high-speed uncapped fixed data connections at home. At the same time, smartphones, tablets, smart TVs, media players like Apple TV, and gaming consoles all reached the point that they were both affordable and advanced enough that they could process video. As soon as both were in place, streaming service usage took off in these regions. In terms of the African market, most service providers agree that greater access to broadband data and the qualitative advancement of internet connections through Fibre To The Business (FTTB) and Fibre To The Home (FTTH) is the first and most obvious answer to the rapid proliferation of streaming services across the African continent more generally. This unfortunately does not yet impact the mass market segments, for whom FTTH is either unaffordable or inaccessible. This market segment remains reliant on mobile broadband data, which is expensive for the streaming services or downloading of content.

As affordable uncapped data options expand across the rest of the African continent, it is expected that usage will grow exponentially. As far industry-specific innovations are concerned, enhanced video compression, to enable more efficient use of mobile data, would be a game-changer. As the larger cloud-based services like Amazon, Google Cloud and Microsoft Azure become more accessible and present in the African market, the cost of content storage, content delivery network services and digital content workflows are plummeting, making it more affordable for streaming services to access the market. For example, the use of LIVE2VOD functionality which essentially converts a linear programme just played out by a TV channel into a VOD asset on the fly. These types of programs can be streamed, downloaded or scheduled for download at off-peak times for consumers to enjoy. In addition, greater adoption of HEVC/H265 streaming protocols will decrease the broadband consumption and improve the quality of streaming services, too. Over the next few years, the quality and popularity of African content will continue to increase, allowing ‘Africans to tell the African story’. User-generated content and new forms of content creation will further add to the impetus and growth of streaming using Kubernetes in Africa.