How The Evolution Of Online Advertising Helped DaniWeb Itself To Evolve

happygeek 3 Tallied Votes 5K Views Share

Dani Horowitz remains as passionate about our online community today as she was back in February 2002 when, while pursuing a computer science degree on Long Island, NY, she founded DaniWeb. Fast forward to now, and an amazing half a billion visitors later, DaniWeb has become a vibrant community of members learning, sharing knowledge, and engaging with other developers, IT pros, technophiles and computer hobbyists. I've been talking to Dani about DaniWeb in the context of a dynamic and evolving online advertising market...

HG: Tell me a little about how DaniWeb itself has adapted to a changing online advertising market?

Dani: DaniWeb was founded in 2002 and by 2003 had become one of the first online 'publications' to independently track and monetize user behavior. Skip ten years and, in 2013, I developed and launched an algorithm to match question askers with likely answerers based on tracking user behavior across DaniWeb. This was a necessary change as by 2013 we were dying a slow death thanks to there being no more demand for Q&A-based tech sites. Between the likes of Stack Overflow, Quora and Reddit they are successfully able to handle Q&As more efficiently than we ever could, and there's no use in competing with them.

HG: So was that a Google thing then?

Dani: It's not that Google is anti-DaniWeb as such, but rather that Google would love us more if we were able to figure out a way to hit their sweet spot like Stack Overflow has done. The problem is that Stack Overflow already has that perfect combination of features and functionality to hit the sweet spot for Google and for the end-user. So you have to design something that is not just great for Google, and not just great for your end-user, but fills a gap in the market as well and solves a need. As a result, I have been focussed on slowly morphing DaniWeb into a social network for techies. The process began back in 2012 when DaniWeb LLC filed the trademark of "Technology Publication meets Social Media". In 2015, when Dazah was launched as a business networking web app integrated into DaniWeb, we took it to the next level. The 2018 rebranding of Dazah as DaniWeb Connect is the next evolutionary step.

HG: So can we look at the role of the online advertising market in all this?

Dani: Twenty years ago, like most forum sites, we made money from banner ads. Today, most of our money comes from members who ask questions paying to be matched and message potential answers, and advertisers paying to message people who fit their target demographic. Our advertising model has changed from banner ads to instant messaging ads. Like the owner of any business, my job is to ensure I can make money and keep that business viable. Thinking of this in terms of an online newspaper helps explain how this works in practise:

One could say that the job of the newspaper business is to increase the number of subscribers and increase the amount of readers. Increasing the number of subscribers directly makes more money, because advertisers pay based on how many readers there are. It is therefore always the case that the amount of money that can be made from advertisers is limited by the amount of inventory available. Rates are set by the market (tech publications always earn X per set of eyeballs; travel sites earn Y and health and fitness sites earn Z) so the limiting factor is always how much traffic there currently is. Perhaps putting more of their best content behind a paywall will increase the number of subscribers. Perhaps working to improve the UI will make the site appear more trustworthy and increase the number of subscribers. Perhaps ads that give readers a sense of FOMO (Fear of Missing Out) will increase their interest in subscribing for fear of not having access to everything and therefore increase the number of subscribers.

To increase the number of readers who visit daily, SEO techniques can be used to gain more traffic from Google. Data mining existing users and providing machine learning to present existing readers with articles they are more likely to be interested in based on their tastes will make the website more "sticky", meaning people will spend longer on it and return more often. Just a 5% increase in return visitors can go a super duper long way to increase the daily readership.

Now let's talk about DaniWeb. Just like every other ad-based publication out there, my business is in selling advertising. The amount of inventory that I have available to sell is limited by how much traffic the website gets. For nearly all of the past 20 years we have had an average of about 5-10 ad campaigns each quarter. The amount of work and effort that it takes to go live with an ad campaign is typically consistent. The variable to how much money can be made is how much inventory we have available. If we don't have a lot of inventory available to sell, each of those 5 ad campaigns can be for $500 each. If we have a lot of traffic, we have a lot of eyeballs to sell, and each campaign can be for $50,000 each. How can I generate that inventory? Well I do that by trying to create a platform that is designed to attract and retain as many eyeballs as possible that meet a certain criteria: the demographic that my advertisers are wanting to pay to get in front of.

HG: So what does improving the platform, and so increasing inventory, actually look like for DaniWeb?

Dani: It involves designing a system that is designed to get users to spend as much time on it as possible, and give them a positive, enjoyable experience so that they will keep coming back. That could be by using a reputation or rewards system to give a warm, fuzzy feeling to answerers. Or to use techniques that bring more instant gratification to askers. Let's take a look at Twitter, for example, Twitter's platform is just a very simple feed of 160 characters or less. As a business, twitter designed a platform to be so addictive that they gained a LOT of traffic (aka inventory) to sell. Tweeters are not customers of Twitter, because they aren't the ones buying anything from Twitter. The same with Facebook, LinkedIn, Google, and every other social platform out there. When you conduct a Google search, you aren't Google's customer. When you buy something from Google, that's when you are a customer.

HG: And when did the online advertising market start to transform, what was the catalyst for you to do the same with DaniWeb?

Dani: That would be in the mid-2000s when Google came out with a completely automated platform to sell ads. It let media buyers buy ads not just from Google, but from every publisher that is in their AdSense network. This completely turned the advertising industry on its head! By making the entire process a self-service platform for media buyers to use it eliminated all of the back and forth overhead between the media buyer and the publication. Campaigns went from taking weeks of back and forth communication to set up, to taking only moments. Not only that, but with all the manual legwork now automated, the process was a lot more efficient and took a lot less manpower. It also brought instant gratification to the world of advertising for the first time ever. Campaigns could now be set up and monitored in realtime. THIS WAS HUGE! Advertisers could now promote products and services that were time sensitive.

HG: Can you explain what you mean by that?

Dani: Let's say eBay knows that you were recently shoe shopping for red sneakers. Previously, the best they could do is show you a targeted ad that might take you to a search page of whatever shoe auctions they currently have going on. However, because campaigns could take weeks to set up, it's possible that by the time you saw the ad there were no active auctions for red sneakers any more. However, with Google AdWords they could now show you targeted ads for current shoe auctions where the ads themselves say "Auction ends in 15 minutes! Hurry and bid now!" And it would take you directly to the auction for that product.

By allowing smaller publications to band together also changed things. When a publication doesn't have a lot of inventory available, advertisers don't want to work with them, because it takes a LOT of manual effort to set up a campaign. Advertisers would rather do the paperwork ONCE and reach all 5 million visitors through one insertion order, than have 500 campaigns going at once totaling the same overall budget, because that's 500 times the work. By working with Google AdWords, an advertiser can advertise across all 500 publications that are in the AdSense network but with one insertion order.

HG: Sounds like a win-win so far!

Dani: Well, obviously this brought a lot of benefits but it brought a lot of problems for publishers as well. It was absolutely great for smaller publishers who previously didn't have enough inventory to land any campaigns on their own, because the most they could sell was a few thousand bucks, and this wasn't worth an ad agency's time, but with Google, they could now enjoy a piece of the pie, and this was great for 99.9% of publications on the web. However, for the 0.1% of publications out there that had enough traffic and were already doing custom deals with ad agencies, this was the worst thing that could happen. DaniWeb was one of those publications.

Google turned ad inventory into a commodity. Before AdWords, setting up campaigns was a long process, and the publisher provided a custom tailored service to media buyers. Reps wined and dined the media buyers. Media buyers bought campaigns based on which publications treated them the best, just like any service-oriented business, it was all about customer service. But then agencies caught on to how they only had to use one platform - Google - and the process was entirely automated, and they could buy their entire marketing budget all at once. This was easier and far less costly!! Publications now had no way of standing out, no way of offering superior customer service, and no way of charging more for better service. Not only that, but because agencies were now JUST buying from Google, even the largest publications were now just grouped together with the smallest guys. Banner advertising became nothing more than a commodity, and with that, the average cost per pair of eyeballs went down 100-fold.

HG: So how did DaniWeb react to this change?

Dani: As a business we had to find alternative ways of making money, because banner advertising wasn't cutting it anymore. We lost a lot of our traffic to Q&A sites like Stack Overflow, so our quantity of inventory went down and the value of the inventory we did have became commoditized. We did what every other big publication started doing, which was explore sponsored emails and then sponsored member subscriptions. The benefit to paid member subscriptions is that you're turning your readers into your customers. Which brings me back to the main points of what lead us to where we are today, why I'm pursuing the path I am with DaniWeb. I can create a MUCH better product and platform if the people using the platform are my customers. This way, I don't have to split focus between appealing to my customers and appealing to my visitors. I'm doing this by charging askers to be matched with potential answerers, which we've been doing with some success now over the past 4-5 years. But I'm breaking out from being a Q&A site and DaniWeb has been moving towards more of a social network model for nearly 10 years now. Which brings me to figuring out additional revenue models.

HG: And they are?

Dani: What can I sell to advertisers who want to get in front of eyeballs if banner advertising is commoditized and isn't going to cut it anymore? I took a page out of LinkedIn's book and decided that selling instant messages would fit the bill. Over the past few years, LinkedIn, Twitter, and Facebook all began using an automated self-serve platform just like Google's. That's pretty much de facto nowadays. It's my goal to turn DaniWeb into a peer-to-peer chat system in effect. I think that we've been successful in that nearly everyone who becomes one of our answerers or one of our more senior members did so because they initially found DaniWeb because they needed help and began by asking a question once upon a time, and they got the answer to their question, and then over time, as they learned more, they wanted to give back. I've spent 20 years as a platform creator, so what I credit myself on is the ability to create a platform that people choose to use over any other platform out there. Right now, askers don't come back on their own. They come back only when they currently have a problem they're unable to solve on their own and Google sends them here. If we could turn this into all members coming to socialize with their peers, we would not be as reliant on Google anymore.

HG: What makes you sure this will work?

Dani: Ever since about 2006, DaniWeb has held monthly meet ups in our office. Techies and programmers often came with a single question they needed help with, but always ended up staying through the night. We went from meeting once a month, from 6-8 pm, to meeting twice a month, from 6 pm to midnight or later. They always came because they were a fresh comp sci graduate who wanted help figuring out what field to go into, or they were building an app and were stuck on figuring out a database problem and thought maybe someone could help, or some other reason, and they ended up making their best friends from the meetup. I can't tell you how many people ended up being a best man at a wedding, or went on to start dating, or went into business together, as a result of our meet ups over the years.

I'm trying to replicate this model in the online world, because, at least for me, I see a huge need. I want to meet people who are fellow tech entrepreneurs and the new platform allows me to do that, and build friendships by finding other tech entrepreneurs who live just a few miles from me who also work from home. I don't know any other platforms that are designed for that. Couple that with this being a new revenue model for us, and I see it as our salvation. We break from Google, get a new revenue model, and make new friends, all at once. I’ve successfully patented the Dazah algorithm, currently the only algorithm patented in the United States that matches people in a social platform based on their chat behavior.

Daniel11 -4 Engineer in Computer

Good to know about the history of DaniWeb and how it evolved over the times to remain competitive.

Marketerector 0 Newbie Poster

Good blog! I truly love how it is nice on my eyes it is. I am wondering how I might be notified when a new post has been made. I have subscribed to your feed which may do the trick? Have a nice day!


Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, learning, and sharing knowledge.