Google analytics – zero visitors but 30,000 pageviews?

Surely something wrong here – look at the following graphs: circled in red are the visitors and pageviews for Monday – how have we got 30,000 pageviews with zero visitors?

Edit: Ok – this is me getting too keen to see the data before it is ready.  Apparently, the visitors number is updated less frequently than the pageviews number so it is possible that visitors hasn’t been updated at all for that day, but the pageviews has. So, if this is true, later on today I should see the visitors number climb upto normal levels…  Funny, I’ve been using GA for years and this is the first time I’ve noticed this.

Align strategy to organisation

All companies fall into line along an axis where at one extreme clear strategy is developed and at the other end no semblance of strategy can be found.  The first reaction is usually that it’s better to be at the end with a clear strategy.  Makes sense right?  Get all your brightest people together and thrash out that differentiating, revenue winning, cost effective, visionary, IP-rich all-in-all-damn-good-strategy.

I’m pretty sure this isn’t the whole answer.  There must be organisations everywhere doing this – boardrooms stuffed with execs ’strategising’.  So what is more important than the strategy?  You know the answer - The Strategy is worth nothing but hot air unless it can be executed.  And who executes the strategy?  The organisation.  Every single person.  If every person within that organisation doesn’t have a explicit link to the company strategy, then you don’t have control.  Consider this: you have this company with six departments.  Each department has a head and a bunch of staff – imagine a road-cone.  It should be just like a real cone – you pick it up by the top and the rest of it comes up as well in unison.  You move it across 1m and cunningly, the whole cone moves.  An executive team are those that set the strategy and should be able to pick up all six departmental cones with ease and move them in the direction of The Strategy.  The reason moving these cones about is easy has nothing to do with the strategy, but all to do with the fact that from the top down, each department is explicitly linked to whatever the strategic line is.  Each head of department has a grasp on their people such that when the company strategy shifts, the entire department can shift with it, just like the cone.

So what holds the department together – what is the glue?  Basically it’s the roles, responsibilities and objectives you give people.  But that’s nothing new.  Everyone does that.  The trick to achieving the cone is, again, obvious:  the role, responsibility and objective of that department must first be defined at the topmost level, before breaking into more detailed chunks and cascaded downwards through the ranks until every person in that department has their little piece to focus on.  This nugget of strategy holds great importance – it demonstrates clearly to the employee what their role is in the company, which particular strand of strategy they are contributing to and a ‘thing’ by which to measure their performance.  If that person performs and meets their objectives (of course the company does everything it can to incentivise them to do so), then that is one small step on the way to executing the overall strategy in a controlled and predictable manner.  In this model every manager in the department has objectives that are the sum of their subordinates’ objectives - all the way to the top, giving rise to this inexorable pressure pushing upwards to ensure that all objectives are fulfilled and hence, doing it’s job in executing strategy.

All companies would do well to be responsive and reactive to shifts in their industry or new ideas such that the executive team can bring about change quickly and as painlessly as possible.  This cascade method of distributing strategy is one way to achieve that.  Tweak a bit of the strategy at the top, and let the changes cascade downwards.  Focus automatically shifts, and the organisation understands what is going on.  I’ve been working on a model like this, and time will tell whether it is successful or not, but as I recall from one Harvard Business Review paper I read once, it isn’t so much the strategy that matters, it is whether you can execute that strategy effectively.  As soon as you can execute a strategy you have the tools to take any strategy and execute it – the glorious things about strategies is that you can dream them up in one afternoon – getting it done is an entirely different thing, and on another scale of difficulty entirely.

Google Analytics goal conversion reported inconsistently

If you look at the screenshot of the Goal Funnel, you can clearly see that the % conversion on the graph (Goal4 conversion rate) is different (about half the value) to the % conversion detailed below in the funnel. I have constrained this to just one day of data for simplicity. There is something wrong here I think. Our Omniture reports are consistent with the funnel conversion figure of 79% – I’ve no idea where the 36.9% figure comes from…

[edit] Ok, so having posted on the GA forum, I’ve been informed that the chart is reporting the percentage of visitors that reached the end goal, regardless of whether they met the funnel entry criteria.  I don’t think the page is clear – it is a page reporting on the goal funnel so I didn’t expect site-wide metrics in there…

SEOmoz community is a great source of information

It’s a great idea. An SEO company can only really scale by adding more people and training them to be good at SEO. Sure, by building systems to assist SEO they can scale further, train faster and deliver sooner but the business is still constrained by people.

The great idea is what SEOmoz have done.

SEOmoz are a company based in Seattle – something that doesn’t matter a jot to me (here in the UK) because I’ll probably never meet these guys or even talk to them on the phone. But they’ll help me enormously SEO-wise because they have a great little add-on to their business – the SEOmoz community. They dish out hints and tips as if they don’t even want any paid business. But of course they don’t worry about that because they can’t hope to satisfy the worlds demand for SEO in an industry that relies on consultants talking to customers. So they do the premium content thing. SEO is a black art stuffed with secrets and for a bargain $49 per month you can gain access to these mysteries. You also get to ask the staff questions *and* you get sensible answers.

They’ve released a bunch of great SEO tools for the general public and hold a few back for the premium members. All the freebies they give out reassure you that they know what they are talking about, and the sheer number of community members (>30,000) tells you they must be doing something right. So when they tease you with the ‘premium membership’ link, you wonder what is there and because you understand the importance of search, you’ll take it.

www.seomoz.org

(I don’t work for them, and they haven’t paid me to write this!)

Rubbish blogger

Man I’m a rubbish blogger.  I read blogs haphazardly, occasionally find real gems then browse away and forget about them.  My own attempts at blogging have hardly progressed past opening up the odd account…  Maybe this year I’ll contribute more…

Database mirroring and hardware upgrades

One of the new features that arrived with SQL Server 2005 SP1 was Database Mirroring. This protects your database against physical disaster (but possibly not logical disaster) and is great for peace of mind. You don’t need to buy another license for the ‘mirror’ server and you can use SQL Server 2005 Express on your ‘witness’ box.

A huge upside of using mirroring is how easy it is when you need to do maintenance on your db server, or you have to upgrade the hardware or swap out the server entirely.

Lets consider a server swap out. You’ve bought a new server, nice dual proc quad core machine with a ton of RAM that you think will keep you going for a bit. It has SQL Server 2005 installed, with the latest service packs and hotfixes applied. The full backup of your old database (which is still in production) has been restored onto it, and you’ve just finished replaying all the available transaction logs to bring it almost up to sync with the production database. Now you’re pretty much ready to send it live.

Just stop the current mirroring session, and run through the mirroring wizard to reconfigure the setup so you have your new server as the ‘mirror’ server. A minute later, you hit the ’start mirroring’ button on the databases you’ve configured for mirroring, and the databases sync up. Once they are synchronised (pretty quick unless they are *huge* databases) you just manually fail over the mirroring so that the new server becomes the new principal server.

Now you have your brand new server running your database in production, with no data loss and only seconds of downtime whilst the databases failed over. The old principal server is now your mirror – do what you want with it.

Temporary tables vs table variables

Query performance can be affected hugely by choice of temporary tables or table variables. Both have upsides and downsides. Faced with either type of table SQL Server will try to keep operations in memory, and both use the tempdb database.
The main differences are: temporary tables are written to the transaction log (so operations can be rolled back), you can create indexes on temporary tables and as such they often perform better when there are many records in the temp table and the query involves joins. Temporary tables however need more careful cleanup – the scope isn’t restricted to the procedure as with table variables. Finally, procedures with temporary tables may need to be recompiled frequently whereas procs with table variables can be pre-compiled.

If the recordsets for the tables are small, and the proc is executed very frequently then table variables might be the way to go. Try both flavours – certainly don’t assume that one is better than the other.

Whoops

Surely the folk at Adobe know how to use the cferror tag!

Comparing a couple of MySQL > MS SQL Server migrators

I have a mediumish sized MySQL database that needs porting to MS SQL Server 2005. The easiest thing to do I thought was check out the many migrator/converter apps there are out there. As they were fairly cheap at around $100 or so, I bought a couple: one from Intelligent Converters and the other from Spectral Core. The MySQL db is about 1GB so not that big, and hardware is a dual Xeon proc machine with 6GB RAM running 64bit Windows 2003 Server.

Tried the one from Intelligent Converters first – it went through the kind of screens you’d expect – pick a source, pick a destination, pick the tables you want to convert and then go. As this one started chugging through it displayed how many rows of which table it had managed to migrate. After 5 mins it was clear that it was going to take some time. Two days later, it had given up after migrating perhaps half the database.

So then the Spectral Core product. This was slightly more expensive, went for the enterprise version that made use of bulk copy so that we’d get better transfer speed. Very similar interface/process – then hit go. 2 mins and 15 seconds later, the entire database was copied perfectly. Happy with that.

http://www.spectralcore.com/fullconvert/index.php