How An Ad-funded Online News Business Works

The success formula for an ad-funded online news business has more to do with savvy execution than quality content. For once, content is not King and this financial model shows why.


Over the last 5 years a range of online-only news businesses have emerged to challenge traditional news publications. Without the heavy costs of a legacy print operation, these upstarts should in theory be able to turn a profit if they can attract enough readers.

The promise of profits in the ‘new’ online news business explains why sites like Huffington Post, Gawker and Business Insider have emerged.

The apparent success of these sites has caused some branded news publications - whose websites were going nowhere - to copy these new models. The UK’s Mail Online would be a good example of this.

More recently, we have seen launches of news sites like Quartz, Re/Code, Vox and The Information (a premium model). So it was all looking good - until GigaOm went bust.

And suddenly online news looked risky again.

To understand more about why GigaOm failed and why sites like Business Insider appear to be going from strength to strength, I’ve taken a look at how an ad-funded online news business works by preparing a financial model and break-even analysis for a hypothetical start-up online news business.

To get started we first need to decide on the content proposition.

The content offered by an ad-funded online news site must have generic appeal. By ‘generic appeal’ we mean that the content can attract an audience in the region of millions of unique visitors per month - essentially, the higher the better.

This means that a niche content proposition is fundamentally the wrong way to go for an ad-funded model. This will become clear later but for now we just have to assume that the content being produced by our news site will be interesting to a large audience.

The next key assumption regards the cost structure of the writing staff. There is a fundamental difference between producing ‘top-quality’ classic journalism and producing a high volume of content that can successfully attract a large online audience.

For example, a story on some aspect of Apple’s plans for the automotive sector could involve a senior level, highly experienced journalist working for weeks on a story. The journalist would identify sources, approach those sources, conduct interviews, engage in related research and fact checking activities and maybe travel to meet interviewees. All before a single word has been written.

An alternate approach would be to search the web for material that already exists in the subject area and then aggregate that material into a story. Perhaps, to make the story look more original the writer could introduce some of their own analysis and supplement the content which has been taken from existing stories, with some other content that the writer has sourced themselves.

Clearly, the cost of producing these two stories is radically different - even though they are focused on the exact same topic.

Because aggregate traffic volume is a primary objective, our start-up news site will have to compromise quality in order to achieve volume.

Therefore, in the following table, I have put together a writing staff that has a skill level commensurate with the production of a high volume of ‘generic’ news content. I’ve also cut costs to the minimum by using interns to help with the writing and using relatively low paid developers to create and operate the website. I’ve also added some resource for SEO/social media and some resources to monetise the ad inventory that will be generated by the website.

As shown in the following table the resulting company comprises 25 individuals, 19 of which are involved full-time in content production.

The assumptions listed below indicate a total operating cost of about $1 million per month, or 12 million per year - 57% of which is accounted for by the writing staff.

The next most important assumption concerns the amount of stories that the writing staff are producing per day, per week and per month.

The objective is to produce a high volume of ‘acceptable quality’ content that has generic appeal for the minimum cost.

Taking into account the fact that interns will not be as productive as experienced writers, the total writing staff of 19 individuals looks more like 14 in practical terms.

If we assume an average production time of four hours per finished story (including editing functions) then this team could produce an average of 27 stories per day (or 133 per week or 572 per month).

These figures have been benchmarked against what the writing staffs of sites like Business Insider are actually producing.

Clearly, we are talking about building a content production machine - rather than a department which excels at purist journalism.

The plan is to source ideas for stories on the web and then allocate those ideas to the writing staff who then produce stories quickly and efficiently, by searching the web and re-purposing existing content with minimal original input.

This is admittedly an extremely aggressive model but as we will see, it is the one which is most appropriate for a profitable ad-funded online news business.

We are now in a position to carry out a breakeven analysis which is shown in the following table, where the key assumptions are highlighted in red.

The total page views of the entire site is around 6 million per month, which roughly translates to 1 million monthly unique visitors. While this is a high level of traffic, it is small compared with the traffic volumes being attracted by news websites like Business Insider (in excess of 30 million per month) and even GigaOm, which claimed over 4 million monthly unique visitors before its abrupt demise.

Another key assumption is that each story should be divided across a number of different pages, each of which carries ad units. Although this tactic can be irritating for users, it will have the effect of dramatically increasing the total ad inventory created by the site by deciding to paginate each story.

We are also assuming an average of six ad units per page. For this simple analysis we are assuming that the website is monetised purely using display ads. This assumption would not be correct in practice because some of the website’s inventory would be filled with PPC ad units (e.g. Taboola, Outbrain, Google etc.). However, the total revenue that these ad units are capable of generating is not dissimilar to what is possible to achieve using display ads.

We can see that our assumptions will generate sufficient revenue to recover our fixed costs completely.

Clearly, it will take a start-up news site years to achieve a traffic volume of 1 million monthly unique visitors. If we optimistically assume that the site takes five years to get to this level with almost no revenues for the first two years then the site would need about $3-4 million in venture capital to reach breakeven, depending on how fast the cost base was ramped-up in the initial build phase.

Although this analysis is based on generating all revenue from selling advertising, it is interesting to consider how the model would be affected if we switched from advertising to subscriptions.

If we assumed an average annual subscription cost of say $500 then a monthly operating cost of $1 million would mean that we would have to sell about 24,000 subscriptions - just to cover costs. This would be completely unfeasible for start-up news site.

However, it gets worse: it is extremely unlikely that anyone would pay to subscribe to the sort of generic content that we have described. The only way to sell subscription-based access would be to change the content proposition from generic content to high quality, original and breaking news content that was not available elsewhere.

This would have two implications: firstly the addressable market would be smaller, which would reduce advertising revenue and the cost of the writing staff would increase dramatically (i.e. probably double).

This suggests that the online news business as it stands today is really divided into two very different business models:

  • Ad supported: This model is based on content that has generic and preferably mass appeal. The content is created using a low-cost writing staff who mainly re-purpose existing stories and content.
     
  • Subscription: This model is based on content that appeals to a niche audience. The content is created using a top-quality, experienced writing staff who produce original stories, source breaking news and are highly regarded as journalists.

Having thought about online news in these terms it is perhaps easier to see where GigaOm may have gone wrong: it seems the site was trying a hybrid model that involved producing high-quality content using an ad supported model - which is kind of a no-mans land. While GigaOm was also in the events business and had a pseudo-research business, these should have been used to increase profits - not to prop up a failed new model.

With such a huge amount of content available online and with the online advertising market as competitive as it is, I just do not think there is enough ‘fat’ available in the market for any news site to do anything other than execute perfectly. Clearly, the starting point for perfect execution is to decide very clearly which of the above two business models you are going to use and don’t fudge it!

(Let me know if you want a copy of the spreadsheet model, which will allow you to change the assumptions).