When I launched 3NJ back in June, I always figured on giving it six months before passing judgement. However, since six months from June 24th would be Christmas Eve, that’s perhaps less practical. And in any case, I’m not sure the full six is really justified.

This is going to read like an obituary for 3NJ, and maybe it is, but I’m not going to absolutely definitely pull the plug on it until I’m done working on the current book (in x days/weeks).

Launch saw 9,000 visitors on the first day. By the end of the first month, 3NJ was down to 100-200 uniques per day. A month after that, 50-100, where it’s stayed ever since, with one small bump when the site got a mention in the Boston Globe.

This is not good.

Similarly, the number of live entries on the site was 468 after the first month. Three and a bit months later, that number has risen by about a hundred.

This is not good.

That’s the bald numbers-as-reasoning, but let’s get into a proper discussion about decisions taken, issues I’ve found running the site and that others have had using it because that’ll be a lot more nuanced and, y’know, useful.

(On which note, answers and comments can be directed at me via Twitter or email. (I haven’t gotten around to sorting per-post Disqus on this thing yet. Too busy with work.) I thought about running some kind of survey for feedback, but realised that those I most wanted feedback from probably wouldn’t be reading this anyway.)

No Names, No Content

3NJ went live with only ten entries on show. Did this discourage people from coming back after checking the initial offerings? I don’t know, but I doubt it, to be honest. The amount of submissions in the first few hours was astronomical (compared to expectations at least), and what I saw of word of mouth back then was largely along the lines of “Cool! This’ll be a place worth checking!” rather than “Why is there nothing here?” The only sane alternative would’ve been to do a broader private beta with, say, 20-30 writers on board to give a wider spread at go time.

(Interestingly, on the day 3NJ went live I was emailed by a staff member for another new excerpt-based discovery site to tell me that two samples of VDZ were up, would I like to pad out my profile or upload more? Oh, and if I didn’t want to give permission to host the excerpts they’d take them down. I eventually ascertained that they’d hand-plucked sampled by skimming free books on Amazon because “you can’t launch a site without content”, and when I explained that asking permission after the fact was getting things wildly arse-backwards they conceded and changed their policy. A spectacular lack of thought rather than anything malicious.)

Open to all, used only by some

The only big disappointment I have about 3NJ’s content profile is that it quickly became an indie (a term I don’t like much but which I hereby concede defeat over) site. Indie writing is fine - hello, plenty of skin in that game here - and it was always going to have a hefty slant that way through sheer weight of numbers (not to mention inclination to try new ways to gain readers) but I really, really wanted this to cover all points along the spectrum. A couple of small digital publishers jumped on - and bless ‘em for doing so - but off the top of my head, and without scouring every ‘about this book’ link, I think only one writer has posted samples from print-published (POD-printed via Createspace/Lulu/etc. excepted) books. Certainly only one from books published by bigger publishers.

I saw a fair amount of talk from published writers expressing interest in it - and had a slightly bizarre but supportive email from one of the Big Five asking if they could remote-dump their entire catalogue onto the site if they wanted - but no actual entries. Forgetfulness, uncertainty over the worth of doing so, worries at publisher level about content rights? None of which would be criticism at all; I’m genuinely curious.

(It’s not confined to published writers, mind, nor to those who never used 3NJ; I know from flagging entries with errors that many users never return to the site. Or at least if they do they don’t bother to check to make sure their own submissions are OK.)

I also emailed the publicity arms, which is where in my experience at least most of the ‘promote your book here’ requests come from to authors, of a bunch of publishers, big and small, US and UK, and received a grand total of zero replies. Not so much as a “bugger off, idiot”. Spam filter, a total hash of communication etiquette on my part, the three possibilities in the paragraph above? I don’t know. All I know is, for all the appreciative noise, 3NJ has had all the actual draw for the traditionally-published as a week-old herring down the back of a radiator.

This has hit both the perception of the site - while indie writing has shaken some of its poor reputation over the past couple of years, it’s still looked-down on by many readers - and, in some cases, the standard of what’s on offer. Much of the content on 3NJ is perfectly fine, but there are (IMO, at least) some clunky entries. That’s fine - my tastes aren’t everyone’s and I always said, and stand by the decision, that filtering for perceived quality wasn’t something I wanted to do and should be up to the reader browsing to do. But the more chance there is of Average Reader hitting a clunky one, and the less chance there is of hitting something outstanding, the easier it is for that reader to dismiss the site out of hand.

Having more published work - where, on the whole, and for all the good examples there are on the indie side, the technical standard is higher on average - on 3NJ would, without a doubt, made it easier to retain readers.

Attention span

Retention is the thing. As well as writers, impressed by the basic concept, saying they’d make use of the site, there were plenty of readers saying the same. That hasn’t happened. Some have come back - I can see returning visitor stats - but most haven’t, not in any significant way. Here I think 3NJ is up against human nature and the mechanics of reading.

Firstly, we all have a comparatively limited capacity for following through on initial impulses over time (see: the amount of money wasted on unused gym memberships). And secondly, as individuals, very few of us read that much. I wouldn’t expect a given reader to need to check 3NJ more than once a month, because who gets stuck for new books that often? There are people who read tons, but they’re in a small minority. For most of us, it’s a book on the bus or the dunny, reading in snatches as we find the time.

To have a regular churnover of visitors, then, a site like 3NJ needs to draw from a very large pool. If the average reader only goes book browsing once a month (he said, making up any old number), to see 1,000 readers a day you need to draw from a pool of 30,000 total. Which is achievable, I’m sure, but hasn’t happened here.

Presentation

The site is deliberately text/CSS styling only. There are no graphics at the front end at all. Obviously that’s part of the concept - but I also chose to do it that way, partly because it makes the site fast (important on mobile devices), and partly because while some assets are readily available for use, I’m not much of a graphic designer. Does the lack of visual sparkle make it less of a memorable place to visit, though? A bit too bland for the ol’ eyeballs to take for long?

There have always been minor issues with the UI and layout, and a couple more fixed in the days right after launch. (There’s a login link, for example, but not a separate register one since registration happens on the login page. No one seems to have had trouble registering - if for no other reason than I linked it carefully at the top of the instructions - but I know it’s something of note.) Again, are these little quirks too much of a turn-off? I don’t know. Tolerances vary.

A dysfunctional back-end

(Something none of us enjoy, fnar.) 3NJ runs on Wordpress because WP is the CMS I know best and could bash something together quickest with. I also knew it could handle a lot of users and deal with an equally large array of content reasonably robustly. And the plugin library for WP is huge and enabled me to include most of the functions I wanted. From the point of view of a submitting writer, I hope the back-end seems to be largely pretty straightforward, clean and clear. I worked hard to bury all but the essentials for the average user and to avoid the cruft that WP offers by default.

However, it is not without problems from the admin standpoint:

• The backstage search functions for both posts and users turn out to be near-useless. You can search with a reasonable degree of confidence for words contained within content, but not within titles (making it impossible to quickly check to see if something’s been posted already; I have to trust to a near-photographic memory and doing things by hand). You can’t search for users at all. It just doesn’t work. Seeing that a user is a spambot and going to find out their IP address for blacklisting currently means going through 40 pages of users.
• Similarly, there’s no way to click on ‘user1’ and see their posts as a list, useful mostly for mass-binning spam but also for double-checking for duplicates.
• Nor can you click on ‘user1’ and ban them and trash their submissions. This is only useful for spambots, but would be very handy - hand-approving submissions means that no spam gets posted, but I have recently begun to have a lot of shite to clear out each time I check.
• Basically, user and post management is very much lacking once you pass a certain size threshold.
• There is no way of providing writers with feedback on their submissions. I have no way of alerting submitters to issues with what they’ve posted. I rely on them checking back and seeing that their entry is tagged as needing changes. This is very sub-optimal.
• The scheduling system works, but not entirely reliably, for no obvious reason at all.
• Login occasionally throws errors, for no obvious reason at all.
• WP’s post ID number assignment is baffling. There are ~600 live entries on 3NJ. The most recent one is numbered 4710. I have no control over it, nor obvious alternative, and it could give people the impression that there was a lot more material than there is.

If I were to try something along 3NJ’s lines in future - which is very possible; I enjoy a technical challenge - I think I’d code it from scratch (or at least from a framework like, say, CodeIgniter). I’d need to get a lot better and more formally organised code-wise than the hodgepodge I’ve picked up down the years currently serving for my technical know-how, but it’s certainly not beyond me. Doing so, of course, means running into more bugs in the early days, but at least the code base is tailored and not something as vast and unwieldy as WP.

Other points

Others elsewhere have raised smaller points when they first learned of 3NJ. I’ll tackle a few here. (Apart from anything else, I’m starting to lose the thread of where I was going and we’re up almost 2,000 words already.)

‘About this book’ links take you to another website and not straight to a ‘buy’ page/don’t take you to Amazon when that’s where I want to get my ebooks from.

There’s no other practical way of doing it, sorry. Hosting files for direct sale would be a logistical nightmare, a rights and accounting disaster in the offing, and so on. Similarly, forcing writers to provide an Amazon link (to which Amazon, pray tell?) is insane unless you’re running a Kindle-specific site, which 3NJ isn’t, while pushing them to an all-markets aggregator like Ganxy is a massive gyp for them and, again, still seems nuts. “Here, run your business this way if you want to be on my site, thankyouverymuch.” If that’s too much of a hoop-jump for readers… well, there’s not much I can do about that.

Send me x anonymised entries to my email every week.

I looked into doing something along those lines, and if 3NJ had taken off I could’ve devoted the time to trying to hack something together, but WP doesn’t make this easy. Coding from the ground up, different story.

Why don’t you pull excerpts from Amazon through their API?

I’m not actually certain that’s possible, not at least without negotiating with Amazon for access to the feed. I’m much more certain that if it was possible, and it was then possible to strip name/author identifiers from material coming down the pipe, and it was then possible to render such content in a manner indistinguishable from direct uploads, that the respective authors might be a bit pissed at having their material displayed anonymously without permission. (See my earlier mention of another site doing this by hand.)

No mobile app?

No. That was certainly beyond me at launch. Not impossible - dealing with a text-only database makes interaction relatively straightforward in theory at least - but a pile of work even if I knew what I was doing. It would have been very useful indeed - the 3NJ approach is ideally suited to mobile reading - but no more than a fantasy.

The end?

So there you have it. 3NJ is probably due to shuffle off this digital coil in the shortly hereafter. It was a fine idea, and still is, I think, and one that was very much worth having a crack at. I regret, to borrow from Piaf, nothing on that score. But this incarnation of that idea never achieved the traction it needed to be worth keeping alive. Maybe future ones will have more success.

I’ll decide for certain in the next few days, but let’s face it, my mind’s pretty much made up.