Business Models for Publishers

Given that I am changing jobs from working for the UK's biggest B2B publishing company (that's RBI, by the way) - I'm on holiday at the moment - to go and work for a Government website that gives advice to Businesses (that's Business Link, by the way) I thought that this post was kind of relevant.  Even more so though now that The Tribune Company is filing for bankruptcy basically because of ailing advertising revenue as an outcome of the credit crunch.  This has a lot of journalists talking - it's their industry and so they are the ones who are trying to work out how to get around it.  In the NewYorker James Surowiecki talked about the future of the online newspaper and over this side of the pond, Roy Greenslade often talks about how bloggers and journalists get along (or not).

This has all led me to this blog post which is to look at the busisness models that publishers adopt (or should adopt) for their online product, along with inherent problems.

Online Advertising

This works in a number of ways:

Display advertising is adding banners, skyscrapers, MPUs, buttons, etc to a web page and then selling them to advertisers.  Traditionally advertisers pay for these at a thousand a go, which probably says something about them.  More importantly, the click through rates (CTR) of these ads have been falling and continue to fall every year.  Whereas advertisers used to think that these were useful ways of driving traffic to their website, they have proved to be unprofitable for direct money making schemes and far more useful for brand making exercises.  There was a concerted approach a few years ago to attempt to get these at a cost per click (cpc) model, but that has fallen by the wayside as publishers looked in horror at persuading people to leave their site for someone elses for a fee.

So if we're saying that they are only profitable from a brand perspective and that it is difficult to prove their worth in a direct fashion (clicks, sales, etc) then we can see why this could be an issue in a credit crunch era.  They will be the first budget cuts because they show very little direct revenue and the return on investment (ROI) appears to be low.

So how do you get around this issue?  Well one way is to make your adverts targeted better, to up CTRs and persuade advertisers that not only do they get brand awareness, but they are getting better direct revenue.  To do this, there needs to be a clear usergroup for the pages which leads us into a classification issue.  If you're a publisher you have to be able to group your pages in a very niche way so that you can sell the adverts to a niche advertiser.  Of course, if you do behavioural targeting eg by using beacons to provide properly targeted adverts for the advertiser you may have some inherent problems in upping CTR.

However this doesn't really solve the problem.  If you then manage to sell all your ads because they are well targeted, your next job is to get more of them.  If you are selling them on a cost per mille (cpm) model then to make more money, you have to persuade the people to load more pages.  How do you persuade them to load more pages?  Well you either get more people to load more pages, or you persuade each person to load more pages.  This leads you to more problems, because if these new people aren't as engaged or if they engaged enough with the content to load more pages, they probably won't have such a high CTR.

This isn't really an issue anyway, because there are lots of sites that rarely sell all their advertising and rely on Google Adsense to get around it (targeted Google ads).  Which does work on a cpc model and hence doesn't really make any money.  If you're not selling all your adverts on the page, then getting more ad impressions doesn't actually make you any more money, so there is no real incentive to grow the site.

This leads me to the next type of online advertising, which is classifieds.  Classifieds come in a number of different formats, but the main one traditionally was job advertising.  Advertisers would pay you a certain amount of money to put an advert of the job that they had on your site.  This follows the old style newspaper system where advertisers could put job adverts in the back pages of the newspaper.  This has its inherent problems too.  Your site has to be targeted enough for the advertiser to think that they are going to get people to apply for your jobs.

This leads you to some KPI issues - you clearly don't want more pages, that doesn't make you any more money.  You don't want any more users, because they won't make you any more money (directly).  Do you want more applications to your jobs?  Well, they don't directly make you any more money either.  What you really want is more jobs being posted, which indirectly means you probably want more applications to persuade advertisers it is worthwhile.  But your KPI really should be 'Jobs posted'.

This leads us to even more problems - we're currently in a recession.  Jobs are being cut left, right and centre.  There aren't going to be more jobs being advertised.  If your business model is reliant on this, then you are reliant on people recruiting and that isn't really happening at the moment.

This leads us to other bits of classifieds.  These can work well as proved by ebay whose entire business model is based on this and this solely, but they can be made part of a site in a broad way (like The Sun) or in a very specific way for that market (see my favourite FWi).  The other trouble is though that there are a series of sites that allow you to do this for free (eg gumtree).

I've seen arguments that suggest this could go either way in a recession.  You might get more people buying and selling or you might get fewer.  I think the principles of the KPIs work in the same way as the jobs though - you need more posts up there, because they make you money.  To get more posts one of the things you need is to have more buyers.

I haven't mentioned any personal classifieds, but they work in a different way making money.  Traditionally they tend to work by making the buyers pay the money rather than the posters, which brings us on to the next topic.


Paid Subscriptions

This is where controversy always comes in.  Traditionally Journalists have written stuff that is then sold on a piece of paper at a newstand.  They always then claimed that if you were going to put it online, you should make people pay for it too.  The trouble is that as soon as one person decides to make it free, everyone else has to as well (why would you pay for something you can get for free elsewhere?).

The only way around this is either to have news that nobody else has (in the B2B market this is fairly common) or information that isn't publicly available.  Here there are a few niche markets - see the chemical industry and most scientific journals - where this works, but it won't work with mainstream news.  Everyone else is publishing the same information, so you have to make it free.

This leads us to a couple of issues though.  Firstly if you are measuring it, you really want to know how many subscribers you have as your main KPI.  How do you get more subscribers is the next question?  This is where the serious issues come.  Either you have to go around the companies and individuals that may want access and attempt to sell them.  This clearly isn't very cost effective (cold calling essentially).  The other option is to tell some other websites to advertise it (eg through banners, skyscrapers and MPUs) or by sending out press releases.

The other way is through search engine optimisation.  Although this too has its inherent problems.  If your content is hidden behind a barrier so that you have to pay to see it, Google can't see it and won't index it.  So you have to give away some for free.  Which starts making your proposition of paying look less attractive, because why would you pay for something that you could get for free?

Sell something

This is the real big one.  Why be a publisher?  Why not sell something on the site.  Traditionally you'd want to sell your newspaper or magazine because you'll do that anyway.  But way just stop there.  Why should you only sell the magazine or newspaper you've been making anyway?  Why not start selling stuff around the niche of your website.  This means that your website probably has to be relatively niche and have things that you can sell in a related way (my best example is the flightglobal shop).  In fact, I don't know why more people don't do this.

Affiliates

This, I feel, should be something that should be relatively easy to do.  Look at the pcworld website where they review products and then link through to places that you can buy them. This is a great idea if you talk about things that you can buy.  And lets be frank, most newspapers do talk about products as do most magazines.  In fact, most fashion magazines have a huge section in them where they put pictures of things you can buy.

So there can be a valid business model, but I think that you need to look at many different areas of money making.  This way you if one of your area fails (they stop paying affiliates or advertising revenues dry up) then you can fall back on your other business models.

Update:

Actually, not sure this really needs to be an update, but I wrote it a couple of days after the original when I realised I'd rushed the end bit.  The other thought I've had is sponsored content.  This sits in two different lines - Display advertising and affiliate deals.  Effectively you are writing content on behalf of someone else (usually an advertiser) with the express point of attempting to pedal some more of their product.  However, you are trying to make it look like you've written it yourself independently.  To maintain your editorial integrity, you are going to tell everyone it is sponsored.  The sponsors will want to put their logo all over it, but make it look like they didn't have anything to do with the content.  So its difficult to do editorially, from a reader's perspective you don't want to be pedalled something from a trusted source and from an advertiser's perspective its hard because unless you get it right, there'll be no ROI.

But if you can get it right, then it can make sense from all perspectives.  The only question from an advertisers perspective is do you measure your ROI only from direct sales or from your increased brand awareness.  And then how do you pay for it - obviously an affiliate deal would be best, but your publisher is unlikely to go for that option, because they'll want guaranteed income.  It's a tough one, but possibly where the future lies.

How to make Slashdot work for you

A couple of weeks ago I wrote a post about how you can make digg work for you (admittedly the hard way, as opposed to the sending links around to a group of mates with a designated time for everyone to digg something).  So this week I decided that I should probably write one about Slashdot.  I think that I'll probably do it in the same style as last time as well (not that you're particularly interested), so that means we'll start with the basics and go into a bit more detail.  As you may well have noticed I have had success on Slashdot before (hey it is a bit of an ego trip, so I mention it as often as possible).

Slashdot
Basic Premise:

Ok - this is a bit more complicated than Digg, but a bit simpler at the same time.

People submit things to Slashdot (these can be logged in or not) - more on that later.

A group of people picked by Slashdot decide what is worthy of going up on the front page.

Everyone else makes lots of comments about it which get voted up or down by moderators (different to the above) and meta moderators.

Posts that don't get put up on the home page are usually lost.

What happens when you register with Slashdot?

When you register with Slashdot you get a whole profile that works completely differently to Digg.  It's really easy to do.  Just by filling in a few details on the create a new profile page, including your moniker.  Be careful with this - your moniker will be the thing that links to your web page if your posts go to the front page.  If you can get you something that is good anchor text then this will be of benefit in the future.  Mine is (rather predictably) whencanistop.

When you create your account you get 'karma'.  This starts at 'neutral', but can change.  If you want the long version of this, you can read slashdot's help.  The basics of it are - the more you comment (when logged in) the more likely people are to moderate your comments up.  The more your comments are moderated up, the higher your karma.  The higher your karma, the more likely you are to get moderator points (meaning that you can moderate people's comments up or down).  Also the more your submissions get posted, the higher your karma gets put to.

Be wary though - your karma may go down as well as up.  If you make comments that aren't in context and get moderated down then your karma may not be so good - this was introduced to stop people having flame wars.  Also anyone moderating you up as 'funny' will not give you extra karma - as they say "You have to be smart, not just a smart-ass."

For a guide on how to get your karma up - read the excellent FAQ section.

I think that having good karma means that you are more likely to get your submissions posted to the front page.  But they claim that this isn't true.

You can also:

• Add people as friends, foes, fans or freaks (these are people who you have added as a friend, people who you have added as a foe, people who have added you as a friend and people who have added you as a foe). You can use this to work out if submitters of comments are people you like or who like you
• You can create a journal in which you can put anything. No really, anything you want. Anyone can read it though, so that might alter what you want to put in there. Vice versa though, you can read anyone's journal. In fact, you can subscribe to it if they are your friend
• You can also submit your journal as a front page post if you wish - this can mean you can create a log of all your entries, even if they don't get on the front page

Just as a note - you can submit without logging in and you can post comments without logging in.  You do tend to get referred to as an 'Anonymous Coward' if you do this.  And anything that gets moderated doesn't count towards your karma.

How do you submit things to Slashdot?

This is surprisingly easy.  Go to the slashdot submission page and fill in the form.  No really it is that simple.  There are a few things that you need to consider though:

• Your username is not important to the story
• Your email or url is not necessarily important to the story (it can be if you want, I suppose)
• The url is the url that you want to specifically link to (you can have more than one though remember)
• Your Subject is the bit that people see as the title of the story and really sells your post.  Make people want to read it on the firehose.  This will mean that to get it posted it has to be good
• The section and topic will determine where it appears on slashdot in addition to the front page.  For example some people limit their pages just to show the subjects they are interested in
• The scoop is the most important thing here.  Make sure you put your link it in.  You can give your link some good quality anchor text and some text that is going to encourage people to comment on the story.  You want to sell it as much as possible here.  This is the bit where you need to have read previous submissions on similar subjects to find out what is going to be good

What if the page has already been submitted to Slashdot?



As far as I am aware - you can't submit a page to slashdot that someone else already has.  So you'll find out fairly quickly.

What is comment moderation and meta moderation



When you submit a comment on Slashdot your comment starts off with a moderation total of 0 (for those that are not logged in or have bad karma) +1 (for thos logged in with ok karma) or +2 (for those logged in with good karma).

If the moderators like your comment, then they will moderate it up by upwards by telling you that you were insightful or interesting.  If they think that your comment was funny, they might moderate it up this way too.  However if they think it is offtopic, troll or flaimbait then they will moderate it down as this.  Moderating something upwards means that it moves up or down one for each moderation.  Many users will only look at comments that have a high moderation score, so the better your comment the more likely people will to read it (and respond to it).

As your karma goes up (with more comments moderated up and stories submitted) you are more likely to get the chance to do some moderation.  If you do get the chance - you will get 5 moderation points (meaning you can moderate five comments up or down by one) and they will last 3 days.  Use them wisely - however you can't comment in discussions you've moderated and vice versa (otherwise you could moderate your own comments up or down).

Meta moderation is something that happens to just the top users.  It is effectively moderating the moderation.  If you moderate badly (ie say something is interesting when it's not, or that it is a troll when it is not) then the meta moderators will pick this up and apply their own moderation.  This could affect your karma too.  Basically you should be careful with your moderation - with great power comes great responsibility.

Is there any spam etiquette?

Obviously everyone is aware that the first rule of spam is don't spam.  But in this case this is not necessarily as true as in others.  If you look at some of the posts up there, you'll notice that the posters always seem to post from the same site.  There is nothing particularly unusual in this - I spend most of my time reading the same half dozen websites that I'd think I'd consider posting from (hey this is news for Nerds - you're not going to get anything from Cosmo on here).  I'd be wary about repeatedly submitting things that aren't ever going to make it.

Look at smiv - he has submitted 70 odd stories and most of them appear to be from the bbc - but a large number have been rejected.  Also look at holy_calamity he has had over 180 submissions - but lots of them aren't being posted.  I think the main rule in this case is not to give up hope.

Using Hitwise dashboards

Given that I am about to move jobs, I thought that it would be worthwhile getting a blog post out about a tool which I currently use, but may not do in the new job.  That tool would be Hitwise.  Hitwise is good in one way that they have a very useful Analysts Blog section, where you can read about some of the trends that they have spotted over time.  In fact, you can also follow the UK arm on Twitter (don't forget that you can follow me too at twitter.com/acochrane) where they put up good links to any interesting articles that they find (thanks for last time :)).

So there is no point me sitting here and typing out some analysis about what is going on with Hitwise that would be useful for you to know.  Instead what I thought I'd do is look at a couple of features that we probably don't use as much as we should, but could come in useful at some point in the future.  In fact, you may have noticed that I've done this bit before as well - starting with 6 tips for using Hitwise that was aimed predominantly at editorial staff (well I nicked if from a presentation that I was giving to editorial staff).  Then I went a bit arse about face and wrote about how you could use Hitwise for your sales strategy, your search strategy and finally for your social bookmarking strategy.  Unfortunately I couldn't think of any more strategies to use, so I've gone for a bit more of a practical approach of how you group all that together.

This is where Hitwise's dashboards come in useful.  When I talked (back in the day) about HBX's new dashboard system I said that I hadn't actually found it that useful in the past.  Now I am going to say the same thing about Hitwise's dashboards - I haven't actually used them before.  Mainly because I just haven't got around to it.  I think that they would actually be quite useful and given the previous posts I've done on the subject you can probably guess what I am going to put into them.  For the purposes of now, I'm going to set up one that Sales would like.

First of all, go to your dashboard menu (it's up there in the top left before you get to custom categories) and click on 'New Dashboard'.  That wasn't difficult, was it.  Note here that you have three choices:

• Base it on a template (there are six of these)
• Copy an existing dashboard (more on this later)
• Create a new one from scratch

Personally I created a new one from scratch, but that doesn't stop you taking a templated one and altering it to suit your needs.  Hitwise conveniently takes your existing main site as the site to use in its template, so if you do like one of the pre-canned ones, it will automatically populate with the right data.

As I said though, I chose to create my own one.  So after choosing the name of your dashboard (you'll notice how original I am in the naming of mine), you get to choose what modules you want to add:

There are quite a lot of them, but I've decided to go with the demographic options, so that we can shoot this across to a sales team really easily.  This will tell the sales team in a very quick way what our user base is like and they can take some of those facts to the customers if they so wish (particularly I'm thinking about advertisers of the display or classifieds types here, but it should also work for lead generation as well).

Below you can see that I've added in a few of the modules and I've configured them to show my website (which in this case is www.fwi.co.uk).  These are easy to configure - you just click on the edit button in the top corner when you mouse over the module and then add in the website name when you get there.  However there are two different ways that you can do this.  Firstly you can select the appropriate modules by ticking the select box as you mouse over them (see below):

Then in the top right hand side of the screen there is an option to 'Modify Selected' under the 'More actions' header.  This means that you can select a variety of reports and give them all to the same website in one easy swoop.  This is great if you have industry report as well as website reports on the same dashboard because you can select them independently.  Also if you have side by side comparisons of two sites to show what they look like, then you can also select just the left hand reports to be one website and the right hand to be another.

The third option is that you can modify an entire dashboard.  This is where it is quite useful when you are copying an existing dashboard.  Mainly because you can set it up perfectly for one of your sites and then just copy and paste the report, modifying it to the new website. 

I like this idea of creating some easily customisable dashboards with lots of information in them, mainly because I think we tend to think of Hitwise as being a bit slow to get any information out of.  The reports that we've created in the past are cumbersome to create and tend to result in a lot of downloading of information into excel and then manipulating it into a way that someone would like to see it.  These reports are easily produced once and can be accessed time and again without multiple set up.  They may not give you the precise information that you would like to see, but that is where you can delve down into the data to give you a more complete view.  This is easily done, because mousing over any particular report allows you to just view that report.  In a way, you could use it in the same way that you would a HBX dashboard, as a hub to get to each of your reports to delve deeper into.

Also of use (and I don't know if the guys at Hitwise are aware of this) is that we quite like copying and pasting these graphs into other reports we send out.  This is really useful here because you can grab a load of reports in one go by highlighting it all and ctrl c, ctrl v - needs a bit of formatting, but it is there.

What would be even better is if you could share this dashboard firstly with other Hitwise users (they could then use it for their sites).  Also would be great if you could share a static version for people without Hitwise logins.  This is where someone from Hitwise can comment and tell me you can already and I've missed a trick or that it is something that is in the pipeline.