People love to interact in realtime. Last year saw the burst of Twitter into the mainstream and the phenomenal growth of Facebook. People love to interact in realtime, whether it is with each other or with content.
And now the world is expecting Twitter to announce this week their version of how users will interact with ads in realtime. Given that, we thought this would be a good time to share some of the insights we’ve learned and the challenges that we see us all facing as we tackle the problem of monetizing in a realtime web environment. (For specifics on our RiotWise product for monetizing the realtime web, click here.)
Some really good things we’ve learned:
- Users are open to ads as long as they’re relevant to their realtime experience.
- Advertisers really want to create ads that are relevant to the realtime experience.
- Realtime applications are starting to make serious money through advertising!
Some really hard challenges all of us face:
- Realtime targeting is complex
- Data is everything
- Advertisers need to be taught how to engage in a realtime experience.
OneRiot has been at it since October (beta launch in January), and in April we expect to exceed 1/2 billion ad impressions across our network of realtime apps.
Our impressions are in the stream (e.g. on Twitter apps like UberTwitter and social desktop apps like Digsby) and they are in search (e.g. on OneRiot.com). Recently we’ve started distributing them across the wider web through traditional ad units.
What we’ve discovered is that when the ad is relevant to a trending topic, relevant to what people are talking about on the realtime web, the click thru rate goes through the roof. In other words, if iPad is trending, and we promote the hottest accessories for the iPad, realtime web users love it. Even on mobile, the CTR goes from an industry average of 0.1-0.2% to over 1% and sometimes even higher. We’ve had CTRs as high as 8% when we really nail it.
Realtime targeting is complex
Every realtime ad engine is different, but the nature of the problem is the same. How do you adjust ads in realtime? People’s attention comes in bursts that you can seldom predict, and a hot topic will move millions of people around the web in minutes. How does an advertiser get in front of those people with an ad that resonates right now?
For ads in the stream, we do our realtime targeting in two steps. First we identify the trending topics for each application and audience. Not just “iPad”, but “iPad Sales Hit 300,000 on Day 1? and “iPad Jailbroken”. This gives us anywhere from 2,000 to 3,000 trending topics per day. Users on UberTwitter tend to like certain categories of topics (defined by CTR) and users on Digsby like other categories. Our system uses CTR feedback to determine how to divvy up the topics by application and audience. This gets tuned and refined throughout the day, in realtime, to optimize CTR.
On the advertiser side we ingest content and products that our advertisers like to promote in realtime as they have it. For content, we take it in using our realtime indexing technology within seconds of it being published. E.g. Coca-Cola had a campaign that produced content around the NCAA tournament. We had their content in our index the minute it was published, and as it became relevant to trending topics we pushed it out to the appropriate applications (Digsby users absolutely love sports content). For products, we index all the top selling products on the web and surface them as they relate to trending topics (from iPad accessories to movie tickets toClash of the Titans – bad movie btw). All ads are created as the products and content become available and so they are always ready ahead of time to be matched to our trending topics.
This is especially important for search, where it is impractical for advertisers to create ads and buy keywords after a term has become hot. Not only is it a lot of work for the advertiser because terms are not predictable, but by the time the ad is created, users have already moved on to the next hot topic of the day. OneRiot advertisers don’t buy keywords. That process would be too slow to monetize on the realtime web. They simply let us index their ads and we do all the realtime matching for them.
Using trending topics and pre-indexing of all ads, we’re able to make sure our ads are relevant to the realtime web user. Even before we started optimizing, our network wide CTR was 4 to 5 times what our partners were seeing using traditional ‘static’ ads. Now that we have the data, we are starting to really optimize our ads in realtime.
Data is everything
Having the technology to optimize in realtime doesn’t do much for you unless you have the data to crunch. And you need a lot of data. We’re currently processing about 12 – 15 million impressions a day. Depending on the app we can start to make a real impact on click thru rates. For an app that has 100,000 impressions per day we can improve CTR by 20-30%. For an app with over 1,000,000 impressions/day we can often double the CTR.
That is huge for the application. Not only are you generating up to twice the revenue as before, but the user engagement doubles as well. No app developer wants ads in their product that hurts the user experience, and no user wants an app that has those ads (will the person who invented dating and mortgage ads please stand up?).
User engagement for ads is key for realtime apps because each one is still in the early days of competing for market share. Poor quality or irrelevant ads might generate some cash, but will send your users straight over to your competitor’s product. Apps with ads that are ‘realtime relevant’ will generate revenue without getting in the way of the user experience.
Advertisers need to be taught how to connect with users in realtime
Probably the most exciting thing about the realtime web is how fast advertisers are moving to connect with realtime web users. Social Media has become a core part of every advertiser’s plan in 2010 and it is only getting better every day.
The challenge though is that advertisers need to learn the difference between connecting with users in realtime and their traditional channels. Coca-Cola might have a great campaign for promoting their brands on TV, but use that same campaign on the realtime web and it falls flat. But when Coca-Cola creates a campaign around the NCAA that is built on original content being published throughout the day by their fans, realtime web users love it. And that was only the first attempt. Crispin Porter Bogusky, Coca-Cola’s ad agency and one of our early partners, will take the learnings from this campaign and make the next one even better.
As brands like Coca-Cola start to show success in the realtime web, other advertisers will follow. The result will be better ads for all of us. Ads that bring us closer to the brands we love because they’re speaking in the language we love…Realtime.
Posted by Kimbal Musk CEO
The beauty of the realtime web is that it’s always evolving, constantly paving the way for exciting new tools and resources to be developed. The Guardian is one of the latest to harness the power of social signals and realtime web by launching Zeitgeist, an eye-catching visual record of what stories people are currently engaging with across the Guardian.co.uk network at any moment. By leveraging OneRiot’s API, utilizing our realtime data, and benefiting from our realtime ranking algorithm PulseRank, the Guardian is able to deliver a colorful and enticing visual that updates with the buzziest stories from across its network. Zeitgeist is great tool for content discovery. It not only filters and ranks the best information, but also offers a user-friendly window into the “zeitgeist” of the publication (not to mention driving plenty of page views).
You can learn more about the “Zeitgeist” by heading over to their in-depth blog post. You can also track developments and provide feedback via Twitter. A huge congrats to Meg Pickard , Head of Social Media Development, Dan Catt and the rest of theGuardian social media team who have been working on the project. And as always, we encourage and invite you to use our API to deliver fresh and exceptional ways to consume media. We’re excited to see what you come up with.
Realtime search is all about speed and relevance, and today we saw a great example in the big Apple iPad event.
By looking into the indexes and determining how quickly we had data about the iPad, and by examining how quickly our RiotWise Ad Engine was able to display relevant ads, we could measure how we did against other search engines. Take a look:
Check out the examples below:
The Power and Simplicity of RiotWise
RiotWise is not about keywords, it’s a fresher more organic way to reach millions of realtime web users. Content owners simply associate an RSS feed with OneRiot’s AMS engine and determine if they want to purchase traffic by the click (CPC) or the impression (CPM). The OneRiot Realtime algorithms to do everything else, including ad creation.
republished from blog.oneriot.com.
The links that people share on Twitter are important signals for OneRiot’s realtime search engine. Broadly speaking, the more people share a link to a specific piece of content, and the faster the rate of sharing right now, the higher that content will appear in our search results. (You can read more detail about our ranking algorithm in this white paper.)
After almost a year of working with the team at Twitter and integrating their Search (aka REST) API, we recently started using the Twitter Streaming API and wanted to share with our developer friends and the greater tech community why we’re pumped about it:
1 – Data volume is fantastic
With Twitter’s Streaming API we are seeing almost 2X the data as we were prior. Stream design paradigm is smart – Twitter is now pushing data in realtime as opposed to 3rd party developers asking for it. Twitter’s old REST API could only be maximized by using multiple threads which would cause duplicate tweets and missing data. Our team had to be very diligent to de-dupe tweets and back-track to reduce the number of missed tweets. Not to mention the complexity of multithreaded programming logic. The new streaming API follows a good design pattern allowing the data to flow in realtime without requiring a second thread. This means less complex programming logic, no more duplicate tweets, and a fully maximized data volume set – a huge improvement!
2 – No more pesky HTTP 503 errors from the search API
The new streaming API allows our data feeds less interruption from HTTP 503 errors (“Service Unavailable”). The old API required us to build a special catch-up thread to make sure we didn’t miss any data during outages. This was a timely and expensive problem. Since implementing the Streaming API we haven’t experienced any service availability issues and have eliminated our “catch-up” process.
3 – It’s easy to integrate
The Streaming API is simply easier to write code for. It took us less than two days to fully integrate the new API with a very small learning curve and a barebones system. I should also point out that the Twitter Streaming API is extremely well documented. (To be fair, so was the last one, but it should be noted that they did a great job with this documentation too!)
No matter what programming language you use (Ruby, Pearl or Java) the integration is seamless. Here’s how we integrate the Twitter Streaming API at OneRiot:
Java is our programming language of choice because it’s fast to develop while delivering high performance. We also use HTTPClient library to connect to the Twitter stream. The tweets are returned in JSON which we parse through right as it comes in the stream. (Side note about JSON: we are pleased that Twitter supports JSON since it’s a lightweight protocol that’s quick to download and easy to read but it’s not as bulky as XML. Oh, and it also has less overhead with clearly structured data.) Lastly, we publish tweets using a traditional publisher subscriber model. Since Twitter doesn’t require a server, we have found that the traditional publisher subscriber model is less daunting than Pubsubhubhub which is more complex and has server requirements.
As you can tell we are big fans of the Twitter Streaming API and would highly recommend any 3rd party developers who have not already converted to do so.
Big News Folks: After a hugely successful pilot phase, we are now announcing the general availability of our realtime ad network, RiotWise, for all developers to monetize their mobile apps, desktop clients, social search engines, and other applications in the realtime web space.
- RiotWise ads are contextually relevant in realtime, resulting in high Click Through Rates (CTR). Advertisers include a range of publishers such as entertainment sites, sports networks, and news organizations. Dynamically created ads link to quality content from these publishers that relates directly to a users’ realtime search term or a global trending topic.
- RiotWise ads have been performing at 3-4 times industry standard CTR for ads in realtime web apps, delivering significant revenue to developers.
- Developers have flexibility to display RiotWise ads in a manner that best suits users of their particular application. For example, partners such as Digsby(the wildly popular social messaging client) and ÜberTwitter (the #1 Twitter Client for BlackBerry) are showing ads directly in the realtime stream. Other implementations include more traditional mobile banner placements or the familiar AdSense-style text block.
- OneRiot shares advertiser revenue with developers.
Reprinted from the OneRiot Blog.
2009 was the year that the realtime web proved its value to end consumers. Millions of people became massively engaged with a new way to consume and share information on the web. Now, 2010 is all about monetizing that explosion of attention and activity. In other words, 2010 is all about monetizing the realtime web.
Several new monetization models are emerging. From various implementations of “sponsored tweets” to innovative “flash commerce” approaches driven by the likes of Groupon, to OneRiot’s own RiotWise network.
RiotWise delivers text and content ads through a range of IAB-standard banners and feeds. We are currently serving over one billion ads per month across the realtime web – on social networks, social media sharing sites, realtime chat clients, news and entertainment websites, leading Twitter apps, and search engines. RiotWise ad copy links to fresh articles from paying publishers, or to product purchase pages from e-commerce partners. RiotWise delivers quality impressions for major brands. The ads work because they are “realtime relevant.” In other words, the ad copy (and the content behind any underlying link) is relevant to the audience right now. This realtime relevance is enabled by a combination of our realtime indexing infrastructure and our trending topics engine (you can read more about that here and here).
Although each of the models mentioned above takes a different approach to monetizing the realtime web, they are grounded by one common premise. The model should be clear to users. It should add value to users. It should not confuse users. If users are confused, they will not engage. And that doesn’t help anyone.
That is why we applaud Dick Costolo’s post on the Twitter blog this morning which, essentially, bans paid-tweets that have not come from Twitter’s own “Promoted Tweets” program. While several “paid tweet” companies have emerged with business models that are sensitive to Twitter users and the wider realtime web ecosystem, several more are not. This ends up confusing users, leading to a poor experience. Although Dick’s comments are obviously focused on the Twitter ecosystem, the end result harms everyone in the wider space.
We’ve designed RiotWise to deliver the right ad for right now, across the realtime web. The ads are content ads (i.e. not Tweets) and help developers of realtime web apps monetize with ads that make sense to their users. If you are developing a Twitter application and want to monetize, our ads make perfect sense as a complement to Promoted Tweets. (Remember to review Twitter’s T&Cs about display rules, and talk to us about best practices for optimal performance.) If you are a developer working on the wider realtime web – monetizing a social network, media sharing site, stream reader, news or entertainment site, search engine, etc – you should also use RiotWise to deliver ads that will resonate with your users right now.
We’ve designed RiotWise to deliver the right ad for right now, across the realtime web. The ads are content ads (i.e. not Tweets) and help developers of realtime web apps monetize with ads that make sense to their users.
For a deeper dive into the RiotWise system, you can read our previous post about what we’ve learned by monetizing the realtime web over the last 6 months. If you’re a developer interested in monetizing your app with RiotWise, you can learn more at the OneRiot Developer Network. And if you are a Marketer looking to reach millions of community influencers on the realtime web, you can sign up for our program here.
2010 is about monetizing the realtime web. But doing so in a way that makes sense to users, supports developers, and helps advertisers reach this fast-growing audience with the right message at the right time. That’s what RiotWise is delivering – right now.
|
|
