<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Greg Connell on Medium]]></title>
        <description><![CDATA[Stories by Greg Connell on Medium]]></description>
        <link>https://medium.com/@gregconnell_99573?source=rss-23a44ba39cc4------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*wnMkwP2lS9bD8ibrmabzrg.jpeg</url>
            <title>Stories by Greg Connell on Medium</title>
            <link>https://medium.com/@gregconnell_99573?source=rss-23a44ba39cc4------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 06:28:47 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@gregconnell_99573/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Why Open Source?]]></title>
            <link>https://medium.com/@gregconnell_99573/why-open-source-bf3b78aa0de4?source=rss-23a44ba39cc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/bf3b78aa0de4</guid>
            <category><![CDATA[netflix]]></category>
            <category><![CDATA[doubleclick]]></category>
            <category><![CDATA[ads]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[advertising-technology]]></category>
            <dc:creator><![CDATA[Greg Connell]]></dc:creator>
            <pubDate>Thu, 08 Mar 2018 02:36:41 GMT</pubDate>
            <atom:updated>2018-03-16T06:16:13.714Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Uzv2Ltxqq-iPGqtR76wBwA.jpeg" /></figure><p>In this article I unpack the following points, which help to explain why RED Interactive moved to open-source the framework we use to build our banners:</p><ol><li>Assess the (the “Stake”) cost/value/saleability of our home-grown tools.</li><li>Consider the solution spectrum: “Big Software vs. Agency Frameworks”.</li><li>Tell the “Netflix Story”.</li><li>Bemoan the “Redundant Inefficiency”.</li><li>Confirm “The Real Value” of what one agency can offer over another.</li></ol><h3>1. The Stake:</h3><p>We have built and re-built our <em>entire process</em> THREE TIMES in the 6 years I’ve been at RED Interactive. That includes:</p><ol><li>The App (project management/compiler &amp; presentation builder).</li><li>The Templates (and the template-assembler).</li><li>The Build Framework (the JS libs used to code-up the creative portion of the ads).</li></ol><p>We were mostly Flash when we started, including the “The App”, which was an AIR/python project.</p><p>Today we use a suite of tools including GitHub (API, LFS), Node (Electron, ExpressJS), ES6 (React, Webpack, Babel, Rollup), and a rotating assortment of the many other, awesome open-source JS packages that suit our needs.</p><p>Everyday is a constant cycle of refactor and improvement, which is driven by myriad factors that are largely beyond our control. To support our small team of 20, myself and a rotation of several other developers are able to juggle the up-keep.</p><p>We certainly cannot justify client line-items for these systems. It’s assumed that we have these things already because we are in the business of building ads!</p><p>But, the maintenance is never-ending, so it ultimately gets absorbed elsewhere on the books. As you can imagine (or know from experience), our business heads get glitchy about it — <em>there must be a way to recoup this cost!</em></p><h4>“Can we license it?!”</h4><p>They asked, and we tried to do so with a CMS (Content Management System) “Ad Manager” tool that we had built. This platform had established us as early innovators of Dynamic Content for banners. Ad Manager made it easier, by orders of magnitude, for us to build an ad once, then change the messaging (images &amp; content) over time, via CMS. This would update all sizes, <em>without needing to edit or re-traffick the unit.</em></p><p>What a great tool! Targeting &amp; message-scheduling is crucial for promoting upcoming events in sporting &amp; entertainment, as well as product release campaigns.</p><p>Our ability to do it well won us our long-standing, AOR relationship with ESPN. And we leveraged it for many other clients, like Microsoft, NBC, Paramount, Nike, Netflix, Anheiser-Busch, UFC, Mastercard, Nickelodeon, Marvel, Lego, Vans, and others.</p><p>This was 2010ish — several years before Doubleclick Studio’s “Dynamic Content”. Flashtalking was still doing Flash. MediaMind was not Sizmek, in fact they were still EyeBlaster. Am I dating myself?</p><p>Even after Google could offer dynamic (for rich media prices), we were able to do the same (minus analytics) in a standard placement — usually at a quarter the cost to the advertiser.</p><p>Ad Manager was a corner-stone of our business for years, but it was fundamentally an internal tool. We could share it with loyal clients, but it was with caveats.</p><p>In order to license it, we needed to rebuild it with licensing as an objective; which we did. However, that process took time (for reasons I discuss in the next section). Also, its development, though better funded than anything we had done before, was still second place to client work. Its dev schedule stretched interminably longer than planned, and by its finish, the industry had (again) shifted.</p><h3>2. Big Software vs. Agency Systems:</h3><p>The hard lesson is that technology is still rapidly evolving —the constant change undermines all efforts, great &amp; small, to build on it.</p><p>I have seen many big players, including Google &amp; Adobe, throw gads of capital at solutions of like variety to Ad Manager: Platforms designed to simplify agency’s creative production. They fail with uncanny predictability:</p><ul><li>It takes too long to build and the industry/tech landscape changes, up-ending the market.</li><li>It is myopic in scope, cannot be integrated in real production, or requires such back-bending to integrate that no actual efficiencies are realized.</li><li>Or similarly, it is not flexible or hackable enough. Or it requires special talent to use. Or it is not compatible with client systems, vendor ecosystems, and/or ad-networks.</li><li>Platforms that get past all those hurdles are generally expensive — it’s almost like out-sourcing the work. Costs cut deep into the agency’s margin, and that expense is in addition (still) to the salary of several pay-rolled experts who can train/coordinate the software’s usage.</li></ul><p>It’s ironic: Competent agencies are the production experts. We have prevailed in every challenge repeatedly, campaign after campaign — from scheduling, to budget, to talent, to training, to creative-restrictions, to tech-evolution, to management. The systems (and people) we have put in place are perfectly tailored to our business. Part of that, of course, relies on an influx of new, blustering talent, eager to prove their worth. The cycle works. Despite whatever burn-out we face — we stay in business &amp; we have a formidable tech stack.</p><p>But these resource-driven, no-frills solutions, these ad-hoc systems do not equal software! The hodge-podge requires a very specialized, very experienced work-force; without which, the wheel 1) does not turn or 2) requires a hopeless amount of support and/or 3) is preferred to be rebuilt (abandoned) per the bias of the next leadership group. All of which are big fails for software.</p><p>All that aside, agencies systems are surprisingly capable. I’ve seen many systems over the years, and they are all competitive in their agility and simplicity. They are like…back-yard experiments: Gossip worthy, straight ghetto. Whatever works, whatever buzzes. The lifespan is ever only until the next stunt. Honestly, to shoot for…something like <em>software</em> is…just not good agency business. <em>#stat</em></p><h4>The Underground</h4><p>So, our “from the mean streets” tech churns-and-burns as many banners as it can before it breaks. It gets fixed, and we keep pumpin’ ’em out. Our rag-tag band of artists and developers find themselves in the position of keeping this monstrous contraption flapping its wings. So many interesting projects and challenges! Some very human struggles are realized, heroic &amp; noble feats of skill abound, and tremendous camaraderie wins the day. What hours we spend maintaining the machine is chalked up as Overhead Costs, and the business people gnash their teeth. Amazingly, we’re still in flight; and we are profitable.</p><h3>3. The Crossover: Netflix</h3><p>So that churn-and-burn mentality was our modus operandi for many years, and with our tools, we owned dynamic creative for many years.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*2PPFzkSf0y--D8-sWL1ZGg.png" /></figure><h4>Monet</h4><p>After a year and a half building on their platform, Netflix returned to us. To no one’s surprise, Monet was costing a lot of mo-ney. They were not seeing the hoped-for efficiency-gains. They wanted our assessment of their platform (which can be read about <a href="https://medium.com/@gregconnell_99573/today-we-managed-to-get-12-different-projects-published-to-npm-node-package-manager-46ae4775da4b">here</a>).</p><p>Long story short: The fact that each agency uses its own, home-grown, mish-mash of solutions was making it impossible for a global advertiser like Netflix to re-use any of the the banners at scale.</p><p>Consider what happens when another region needs to make a creative/functional tweak to an existing ad? Well, those local agencies pitch to rebuild the campaign from scratch because “it will be faster than reverse-engineering another agency’s delivered banners”. Makes sense. Netflix tries to pay for just a small change-order, whereas the local agency is looking at a huge deconstruction to figure out what the heck makes this foreign banner work. So the partner incentives are all wrong: Rebuilding is a bigger pay-check for the agency, and using their own tools is easier.</p><p>On top of that, on Netflix’ side, each rebuild was more expensive than before because the dynamic-operability with Monet is way more complicated to implement!</p><p>A unified build-process was needed — one that would be usable by all Netflix partners.</p><h4>Collaboration</h4><p>Netflix knew how much redundant work they were paying for — rebuilding creative just to localize/tweak one that was already done. And Monet integration was expensive. Unless their partners could work with each others’ projects, the cost of a dynamic platform was just…<em>more overhead</em>.</p><p>Working together with several other agencies (including MediaMonks, Circus in Mexico, and CloudNine in Germany), we helped Netflix chart a course that allows for greater collaboration/sharing of components.</p><h3>4. Redundant Inefficiency</h3><p>After getting intimate with some groups that I used to consider “competitors”, I realized something: We’re all solving the same banner-tech mandates:</p><ul><li>Use all available capabilities, mitigate client expectations, and play nice with publisher requirements.</li><li>Retain technical flexibility so that, when a new gimmick is all the rage, it can be integrated into the pipeline.</li><li>Maintain consistency &amp; ensure the product is true to the creative/strategic vision.</li><li>Keep your code-base &amp; work-force healthy, so that you can gracefully transition the major tech epochs.</li></ul><h3>5. The Real Value</h3><p>If you are in the banner business, you (or a couple of your hard-working developers) are maintaining a system that solves those problems. One way or another, you “turn ’em out”.</p><p>Yes, some agencies do it better than others. But our sales team found technical excellence is non-persuasive as a selling point: The client isn’t concerned with our internal efficiency (how the factory works), or whether the code is “right” — it could throw a solid console of errors, it’s doesn’t matter. From a creative agency, the client cares that 1) the price is right, 2) their brand looks good, 3) it clicks through, and 4) it will pass the Ad Network’s QA.</p><p>The restrictive nature of ad real estate puts another nail in the coffin: How much can you distinguish yourself with 200k? Endless restrictions: No 4th party calls. Sandboxed iframe. Unsafe code. Unsupported file types, ancient browser compatibility, etc, etc, <em>ad nauseum (pun intended).</em></p><p>Lastly, there is talent. Restrictions + the poor reputation of banners + endless drudgery of writing disposable code. These job characteristics do not attract the highest caliber talent. It’s the same with the scores of people up &amp; down the chain, to the media agencies all the way to the publishers. How to say it nicely? Ads are…a way to make a living.</p><p>Even when an agency does come up with something clever, it is often crippled by the ecosystem because it asks for a disgruntled middlemen elsewhere in the chain to “take a risk”.</p><p>As further evidence of advertiser ignorance: After Flash died, in the early H5 banner days, we innovated an image-flipping/compression system to achieve auto-play video. We had this for over a year before all browsers supported the &lt;video/&gt; tag in standard placements. To our surprise, clients were not moved. In fact, <em>they did not know they didn’t already have auto-play video</em>. Try to sell somebody something they think they already have!</p><h4>The Real Value Is</h4><ul><li>Creative/strategic thought leadership.</li><li>A well-coordinated traditional, display, and social-media plan.</li><li>Pixel-perfect designs, beautiful animation, good interactions.</li><li>Professionalism: Delivering a quality product, on time.</li></ul><h3>So…Why Open Source??</h3><p>To recap:</p><ul><li>For an ad tech platform to be truly competitive as a licensable package, a bigger stake is required than agencies can risk.</li><li>Large scale ad tech projects tend to fail due to 1) rapidity of change, 2) their distance from the “in the trenches” realities.</li><li>To be in the business, each agency solves the same, hard production problems with ad tech, but they can’t bill for it, and it isn’t a sales point.</li><li>The distinguishing quality of an agency is their strategy, cultural awareness, eye for quality, and professionalism…not their ad factory.</li></ul><p>So…essentially, <strong>why not</strong>?</p><h4>Advantages of Open Source</h4><p>Open Source is one of the modern miracles, as important as the invention of the Internet itself. In a future article, I will more deeply discuss the profound collaborative effect that GitHub (and Linus Torvalds aka “Linux”) has had on the whole world.</p><p>Open Source projects are fluid, intelligent, evolving projects. More than just being publicly available, project consumers are able to submit improvements back to the codebase. This cycle builds momentum &amp; trust, and this exponentially accelerates the rate a project matures. Contemporarily speaking, it’s the only way to build something competitive.</p><p>In other words, like-minded professionals facing the same problems:</p><ul><li>contribute to global efforts, in concert</li><li>keep their code up-to-date with evolving environments</li><li>let their bad assumptions be challenged</li><li>thrive when their applicable value is highlighted</li><li>find confidence in the strength in numbers</li><li>achieve massive momentum</li></ul><p>And all that work happens for free, because frankly, it’s just too damn hard to convince business people of what’s what when it comes to low-level tech. Developers know if something works, and that’s how projects get adopted and gain reputation.</p><h4>Portability</h4><p>Ads are the most pervasive tech in the world — they run on all platforms/devices/os’s. Many groups/agencies/people handle distribution at all phases. A sharable, dev-able, fully capable format is badly needed. One that can keep up with the times.</p><h4>The Right Way</h4><p>In the process open-sourcing, the project discovers its <strong>real</strong> standing in the tech world: It could fill a big vacuum and become very popular, or it could realize itself satellite to a stronger project of the same variety. Both are good outcomes. Popularity leads to opportunities. In the later case, the realization of a stronger initiative helps the original project to realign itself accordingly. They may even try to inherit from it to save themselves the work. In so doing, they become more strongly positioned for the future.</p><h4>Monetization</h4><p>Momentum and visibility beyond a single office yield better opportunities to monetize than ever before.</p><p>There are Service Contracts. Just because the code is free doesn’t mean attention from the core developers is — dev hours that have never been billable potentially become so.</p><p>As well, there are clever ways to position oneself, where the open-source portion is the bait and another, licensed portion is the catch.</p><p>But mostly, if successful, the benefit is that the business converts the expense of maintaining an in-house, complicated tech stack into a global, community effort that 1) increases capabilities and 2) lowers cost.</p><p>Also, it’s a bad-ass story to tell.</p><p>Tech reputation is powerful — it attracts talent, which leads to better direction &amp; a stronger position in the face of change. Also, better position means alignment with bigger forces. From there, real connections are possible. Bigger connections leads to bigger opportunities. Ultimately, the direction of the industry is influenced.</p><p>All of that fits nicely with our case-study and PR goals.</p><h3>FF0000 Ad Tech</h3><p>Realistically, even if our build tools are available for free, adoption will be an uphill battle. Having Netflix push it on their partners will help. Extending it to Droga5 will help.</p><p>But in the wild, agencies who already have a process are invested. They will resist change, even if it’s for the better. Or maybe they have a superior process, which we need to realize, to incorporate &amp; improve our offering.</p><p>In the end, it really does come down to <strong>why not</strong>?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/256/1*gtnuN_Kbm-7KptfQ04m_9g.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bf3b78aa0de4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Technology Valuation]]></title>
            <link>https://medium.com/@gregconnell_99573/daily-technology-valuation-95dee5b9a534?source=rss-23a44ba39cc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/95dee5b9a534</guid>
            <category><![CDATA[advertising-technology]]></category>
            <category><![CDATA[valuation]]></category>
            <category><![CDATA[labor]]></category>
            <category><![CDATA[commit]]></category>
            <category><![CDATA[git]]></category>
            <dc:creator><![CDATA[Greg Connell]]></dc:creator>
            <pubDate>Sat, 23 Dec 2017 06:08:29 GMT</pubDate>
            <atom:updated>2018-01-08T07:56:09.861Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*oLZMeii_AX5_vqrixzABvA.png" /><figcaption>My GitHub Activity in 2017</figcaption></figure><p>Anyone who participates in the Open Source community will recognize this chart — it exists on each individual’s public GitHub profile. Mine lives here <a href="https://github.com/gmcdev">https://github.com/gmcdev</a>.</p><p>The chart shows a duration of time: Seven days per column; each column is a week; the current week is all the way on the right-hand side; going left are the previous 52 weeks.</p><p>Gray dots indicate no “commits” were made (to be unpacked below). Green dot variations represent a quantity of commits. So the solid dark ones are the busiest days.</p><p>All told, it’s an easy visualization of a year’s worth of work.</p><p>Let’s unpack it a little bit more. To do that, we need to know the project’s: 1) context, 2) scale, and 3) state.</p><h4>Project Context</h4><p>This project is a technology endeavor that builds online banners.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/419/1*j0wxYOzEWu-APrr5ZBJuYQ.png" /></figure><p>The system includes 3 major applications, a templating system, and an authoring framework.</p><p>The <strong>applications</strong> are:</p><ul><li>A project management system</li><li>A project compiler</li><li>A presentation builder</li></ul><p>The <strong>templating</strong> system includes the patterns &amp; starting-points for all of our various markets.</p><p>The <strong>framework</strong> is an ecosystem of tools that:</p><ul><li>enable medium-wage workers to do the business</li><li>make the production hours more predictable</li><li>guarantee the operation &amp; quality of the product</li></ul><h4>Project Scale</h4><p>We are a small department, less than 40. The number of people working directly with the tools every day are less than 20.</p><p>Support &amp; maintenance of the system requires several different skillsets. With proper staffing, two higher-wage developers are able cover this.</p><p>Enhancements require an architect (and additional developers, depending on the feature) because the system is complex.</p><h4>Project State</h4><p>The lifespan for this type of system is not more than 5 years. More likely, 2–3 is all you can get before the surrounding environments out-pace you.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*Lap0LLTIF_qcyhRLm5SP6A.png" /></figure><p>We are in year 4, apparently still viable. All parts of system, aside from the original python backend, have been rewritten to stay current.</p><p>Our product’s format, compatibility, and performance cannot yet be questioned by the media agencies and publishers who run our ads.</p><p>However, all signs in the market make it clear that an evolution is, indeed, happening around us.</p><h3>2862 Commits in 2017</h3><p>As you can see, there are several weeks of vacation in May, mid-October, and a few long weekends. At the same time, there were plenty of weekends worked, especially toward the end of the year.</p><p>For the sake of easy math, let’s just count 260 work days across 52 five-day weeks. This yields an average of <strong>11 commits per work day</strong>.</p><h4>What is a Commit?</h4><p>A commit is a group of edits to any number of files in a project. It’s like a checkpoint.</p><p>For the sake of clarity / organization, developers try to organize commits by task or feature completed. Theoretically, a commit is a reasonable, workable snapshot of the code-base.</p><p>The motive for a commit is generally 1) Introduction of a feature, or 2) Addition of a hot-fix.</p><p>Features are improvements, usually necessitated by a realization that the system would work better if such-and-such were possible. The timeline for a feature is usually well-planned and paced — often secondary to more immediate needs.</p><p>Hot-fixes are bugs &amp; issues that have been brought forth by usage. These are generally accompanied by a user who is blocked, which means they are almost always dire in nature and in need of immediate attention.</p><p>Translated to english, then, this means that my daily work-rate is equal to 1.375 features/fixes per hour, every work-day of the year.</p><p>If I worked 8 hours day, every day of the year, that’s 1 commit per hour.</p><p>For 2018 I have vowed to do better at labeling my commits as “feature” or “hotfix”, to get a better understanding of the overall health of the system.</p><p>I would hope that the majority of commits are features not bugs.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=95dee5b9a534" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Netflix Web Components]]></title>
            <link>https://medium.com/@gregconnell_99573/today-we-managed-to-get-12-different-projects-published-to-npm-node-package-manager-46ae4775da4b?source=rss-23a44ba39cc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/46ae4775da4b</guid>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[github]]></category>
            <category><![CDATA[netflix]]></category>
            <category><![CDATA[advertising-technology]]></category>
            <category><![CDATA[npm]]></category>
            <dc:creator><![CDATA[Greg Connell]]></dc:creator>
            <pubDate>Fri, 15 Dec 2017 02:47:30 GMT</pubDate>
            <atom:updated>2017-12-23T06:11:58.235Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TgJPuTO5AFZwc0VPgCGWHA.png" /><figcaption>@netflixadseng on NPM</figcaption></figure><p>Today we managed to get 12 different projects published to NPM (Node Package Manager).</p><p>These projects are Javascript components designed to bridge data from a Netflix CMS called Monet to banners being produced by partner agencies.</p><p>To do this, there is a master component that manages communication with the backend. It manages various types of fallbacks, but in the case of success, it returns promised data to all of the ad’s view components.</p><p>The view components are common things like Video Players, Call-to-Actions, Netflix Ribbon logos, etc. Historically, each agency employed by Netflix would develop some version of these components.</p><p>All of those different versions were making it impossible for Netflix to automate, reuse, or scale their ad production systems.</p><p>With this release, Netflix now has a distribution mechanism that is compatible with all of their global partners. And they have a set of base components to offer…</p><p>This is an awesome accomplishment, at least in the advertising industry (reasons for which are plentiful but are subjects for another day) — suffice to say, only so much investment goes into code that only runs for a several seconds, for a day or two; and whose surface is generally hated/blocked by the majority of internet users.</p><p>However, with Netflix components made available to their partners via NPM / GitHub, they gain the ability to :</p><ol><li>Better control the quality of their ads.</li><li>Push updates/fixes to the function of their ads.</li><li>Avoid paying multiple agencies premiums for 1 &amp; 2.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=46ae4775da4b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[FF0000 Ad Tech]]></title>
            <link>https://medium.com/@gregconnell_99573/ff0000-ad-tech-64341e03ccc0?source=rss-23a44ba39cc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/64341e03ccc0</guid>
            <category><![CDATA[banners]]></category>
            <category><![CDATA[webpack]]></category>
            <category><![CDATA[javascript-development]]></category>
            <category><![CDATA[advertising]]></category>
            <category><![CDATA[es6]]></category>
            <dc:creator><![CDATA[Greg Connell]]></dc:creator>
            <pubDate>Fri, 01 Dec 2017 20:02:00 GMT</pubDate>
            <atom:updated>2017-12-01T20:02:00.941Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*sj9Rr2zbJT8an2DZjzhfyQ.jpeg" /><figcaption>3420 Ocean Park Blvd., Santa Monica CA</figcaption></figure><p>Hello world~ I am the VP of Advertising Technology at RED Interactive in Santa Monica, CA.</p><p>For the last six years of my career, I’ve been building efficiency tools for the creation of display ads. Unlike apps &amp; websites, banners have a much different disposition in terms of budgets, talent, development-time, life-span, and run-environments. The k-weight and capabilities are often heavily restricted.</p><p>That is to say, building HTML/CSS/JS for ads is a rather different beast.</p><p>At RED, we’ve tried to solve these problems with many of the familiar techniques:</p><ol><li>Version Control</li><li>Templating Engines</li><li>Frameworks</li><li>Build Tools</li><li>Content Management Systems</li><li>Performance Analytics</li></ol><p>In 2017, we were engaged by Netflix to help them with their proprietary dynamic advertising platform called Monet.</p><p>This work has stimulated a full modernization effort of our internal tech stack.</p><p>We have decided to open-source our tools and process. This series will document some of that development.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=64341e03ccc0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[FF0000 Ad Tech]]></title>
            <link>https://medium.com/series/ff0000-ad-tech-56444f4676e4?source=rss-23a44ba39cc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/56444f4676e4</guid>
            <dc:creator><![CDATA[Greg Connell]]></dc:creator>
            <pubDate>Fri, 01 Dec 2017 19:56:07 GMT</pubDate>
            <atom:updated>2017-12-01T19:56:07.767Z</atom:updated>
            <content:encoded><![CDATA[<p>Hello world~ I am the VP of Advertising Technology at RED Interactive in Santa Monica, CA.</p><p>For the last six years of my career, I’ve been building efficiency tools for the creation of display ads. Unlike apps &amp; websites, banners have a much different disposition in terms of budgets, talent, development-time, life-span, and run-environments. The k-weight and capabilities are often heavily restricted.</p><p>That is to say, building HTML/CSS/JS for ads is a rather different beast.</p><p>At RED, we’ve tried to solve these problems with many of the familiar techniques:</p><ol><li>Version Control</li><li>Templating Engines</li><li>Frameworks</li><li>Build Tools</li><li>Content Management Systems</li><li>Performance Analytics</li></ol><p>In 2017, we were engaged by Netflix to help them with their proprietary dynamic advertising platform called Monet.</p><p>This work has stimulated a full modernization effort of our internal tech stack.</p><p>We have decided to open-source our tools and process. This series will document some of that development.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=56444f4676e4" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>