We can do it (!/?)

In this article Dave Trott talks about the famous US "can do" attitude and compares it unfavourably to the UK attitude.

In New York, everything I got in trouble for in the UK was just what they wanted me to do.
Also, because America is a younger country it has a clean slate.
This gives it a ‘can do’ attitude.
If it’s a great idea let’s go for it.
As an older country, the UK has a ‘can’t do’ attitude.
This idea’s never been done before.
There must be a reason why no one’s done it.
So we’d better not take a chance.
Also I loved the anti-elitism of America.

I present the case for the defence...

Mobile usage is increasing, haven't you heard?

Another article from Omniture relating to the success of mobile.

Lets consider a few things:

1) Mobile usage is increasing, we all know it. Showing revenue increasing is comforting but not much in the way of news, its what we'd expect as the industry matures and stores provide more mobile commerce facilities.
2) Regarding the % of revenue derived from mobile devices. They mention 0.4% (wooo) - what I'd like to know is if there are any common factors where the increase has been highest - is it dependent on the product/vertical or technology (i.e. these retailers had improved their mobile facilities)?
3) Revenue growth outpacing page view growth, is that surprising? We don't want to be downloading loads of pages on our mobiles, perhaps the revenue growth out gunning page view growth can be explained alone by site improvements i.e. less pages in the purchase process. Although that is possible, I anticipate it is simply more people buying via mobile. Whereas before it was something new and people were browsing around the sites rather than buying, they are now purchasing. We are seeing increasing growth on an activity that was probably not very common in the first place.
4) The Iphone stat. I'd love to know what the % split of revenue is from Iphone apps to site browsing, that's surely one of the more interesting figures.

To find that the increased market share of smart phones in the population relates to increases in one of the activities they were aimed at (ecommerce) is not really much insight.

Its good that we're showing mobile is maturing and becoming an increasingly relevant player as an ecommerce channel, but I could have told you that just from asking my friends about their mobile usage - I don't really need the stats.

What I want is to know more about how to take the greatest advantage of this channel which I think is something missing within the blog post and an opportunity missed to demonstrate Omniture's value of tracking mobile devices - it really is a good idea!

So why should I post about something that isn't news? I just hope that the increase in mobile usage comforts you in this season of goodwill - it confirms what all those expensive trendy market analysts have been telling you would happen.

And lets face it, if you haven't got your site geared up for mobile its probably a bit too late to worry about how to gain that extra 0.3% in christmas revenue.

Happy mobile christmas!

Cool new blog: Rory Sutherland

Here's a chap who I like the sound of.

Rory Sutherland

Decribed by himself on twitter as the "Fat bloke at Ogilvy". That alone makes some interested but he does have some interesting viewpoints, wonderful turn of phrase and great presentation skills.

Take a look for yourself.



Thanks to Grumpy Brit for introducing me.

Death of Search?

So the ex CEO of Omniture has produced a blog post.

Its very interesting that it aims to deflate the search bubble, and lets face it, when we say search bubble, what we really mean is Google. To me its like a shot across the bows, a message of intent if you will.

I have for a long time believed that search is over emphasised as a mechanism for sales and I'd like to focus on this "retail" side of search for this blog. There's been a great number of experts encouraging customers to throw money at search - it is after all very "measurable". This "measurability" has made it easy for customers to present to their board and show how "effective" their campaigns have been. The conversion rates often aren't spectacular but when the CEOs compare this to the woeful effectiveness of banner advertising, its seems attractive. For CEOs, search conversion rates are accepted as being the cost of doing business online - an essential, just like your electricity bills.

However I dont think the competition from the new holy trinity of facebook, twitter and iphone that has convinced me of this.

I've consistently said is that search doesn't "sell" - search "finds".

If I'm looking for a company and cant find it through search this will definitely have impact my likelihood to buy from that company. If I've bought a commodity from a company before and had a good service and price, I will return to them for similar items - but that's not down to their search ranking.

If I'm looking for a commodity, a holiday, a new blu-ray player anything like that, seeing a company at the top of paid search will have some impact but it will be minimal. I'll respect them for having money and some targeting, but this alone will not make me want to buy from them.

What I'll want is a mixture of human advice and product information on that commodity. I'll be looking for information from other people (reviews, ratings etc) and price comparison sites that tell me where would be a good place to get it from.

So the next question I would ask of the trinity is can they provide me with that information?

Lets look at them.

Facebook contains communities. Does Facebook have the potential to become this place that I would go to find reviews? Yes. But its still not yet a place I'd go to for reviews, I'd seek out user forums or independent (or supposedly independent) sites that house reviews from industry experts. User forums / sites / applications are invariably more sophisticated and in depth than anything on facebook, although of course this could change if facebook opens up more.

Twitter. Communities? Yes. Reviews? well not really, how much can you fit into the twitter text box that tells you anything other that "its great lol" and "it sux". Twitter is very much about immediate impact and news (hence why it gets so much media attention). I think it could be good for generating excitement of a new product or company, but this would be short lived as the twitterverse moves on. Twitter can of course link to other sites and this is also of great value. If you can get your reviews on twitter quickly then someone will find them when they search in a week or a months time.

IPhone apps? Communities? Yes, you could create that app. Reviews? Yes you could create that app. But how do you get that app on a users phone? That's the tricky part. You've got to create a compelling reason, and that has less to do with the iphone and more to do with what you offer.

So for me, yes there are alternatives to search that google should be worried about, but nothing that immediately threatens it dominance in the online space. I'd recommend reading the Ad contrarian for more interesting articles around the other "dead" channels that seem to be ignored.

I think there is still space for someone to achieve dominance in independent advice.

I think the "old guys" at amazon and ebay offer more interesting and threatening alternatives than the holy trinity.

I think companies like Yellow Pages and Which? in UK are probably missing this opportunity as well, the ability to build these communities and user reviews whilst maintaining a relative air of independence could be done easily given their branding & audience reach.

And finally, I'm also wondering what happened online to all those governing bodies, organisations and charities that certify and review companies offline in the UK. I imagine they could make quite a name for themselves if they had a larger web presence, the likes of Tripadvisor performs the role for me that actually I'd like to come from ABTA. This bodies could be more than just a record of who agrees to their codes of practice and actually contain direct customer feedback.

Anyway, I digress. Is search dead? No. Will the holy trinity replace it? No, but it will divert some of a companies online resources towards it because the next generation are now churning out social media gurus rather than search guru's because for todays kids its the new way to make a fast buck at their elders expense.

So to finish with an example:

Does anyone really need to "search" for "Apple" or "Iphone" in order to buy it, no. But I bet they would search for "cheap Iphone tariff" or "Iphone review". Do apple need to have a big facebook or Twitter strategy? Probably not, because they've got the product and market share.

And what about everyone else? Everyone with a product that less revolutionary, less loved, less well known and more mediocre? Is Facebook and twitter going to make the difference in their marketing and sales over and above any search activity?

The channel is the mechanism of the message. You have to get that message right and deliver it to your audience in order for it to work. Unless a channel defines your audience you should concentrate on the message first, mechanism later.

My Local Sainsburys

My local Sainsbury's has just gone through a makeover.

They've shifted all my usual purchases and I struggled to find them.

I found them eventually but not until I'd seen a load of stuff I don't normally see (or am normally tempted by).

Surely they'd optimised the layout previously, surely everything was perfect, what have they done?

What they've done is amend the layout for a number of reasons both customer and business focused and whilst hopefully considering the risks of offending regulars. They will of course measure and evaluate the changes but I doubt they'll roll back to the way it was unless the takings take a dramatic turn for the worse, and that's unlikey.

And it actually doesn't make a difference to me, I'm still likely to shop there, I've been exposed to a few different things all for the inconvenience of a slightly longer shop for one day. Sainsbury's get a bit more insight on shopping behaviour, more new views of different products and I assume a better functioning shop. I know for some people, change is a bigger deal than others, but seriously, I wonder how many people would leave because of overall layout as opposed to price, location and products.

I was more upset when they stopped producing their mixed-olive humus. No seriously. That pushed me to a few more shopping trips to Morrisons and Waitrose than this change will.

So why blog on this? Well people are often so scared to make changes to their sites. Yes change can have a big negative impact, but it can also reap rewards. Just make sure when you change the layout you don't forget that what drives most customers are the products and their perceived value.

BRING BACK MIXED OLIVE HUMUS!

Is Web Analytics better than everything else?

In joining Adobe, I've just had to redo some HR related work suggesting that I will also act with integrity and not release trade secrets or intellectual property.

I think this a good thing, but I did wonder how it applies to what I know from being in the on and offline analytics industry for 10 years. I was also asked recently by a colleague whether I thought that online analytics had surpassed applications / methods used offline so I thought I'd use this blog to talk about novelty and the differences of the channels without giving away any trade secrets.

My answer was that although web analytics has developed considerably over the last few years, offline analytics has kept pace and there are still areas where its weak, I thought it might be worth discussing.

Data Collection: This is actually hard to compare. Web Analytics vendors strength is its ability to come up with new ideas about how to collect new data - facebook, set top boxes, mobile phones, twitter, search engines, games consoles etc. all these appear to be in the remit of the "web analytics" vendor that surely sets it aside from offline analytics. If you think that people aren't thinking of new ways to track transactions, phone calls, sewer systems (someones got to!) and various libraries of information held in digital channels from the more traditional vendors, you'd be highly mistaken.

Data volume: Are online analytics vendors storing things in advance unique ways? I don't know. I know that cloud computing is a big thing. I know that Google make use of it for their searches, how much they use in their analytics, I don't know but I'd love to find out. Do other vendors use similarly adventurous methods of managing data volumes? I don't know. I know that Omniture Insight (ex Visual Sciences) does some pretty cool things but is this more impressive that anything "offline"? Vendors for Telcos will pitch to record every single SMS, MMS and phone call interaction, a retailer will record every single transaction made at a point of sale, a bank has to record every time you add money to your account, do you not think that is impressive in terms accuracy as well as volume? I don't care if a hit from my site goes missing, but if that's my paycheck its a very different matter. Not only all that, the offline vendors invariably tie into the operational systems, whereas online, importing operational data appears still in its infancy. The "online" analytics that impress me are the huge retailers, Amazon, Ebay, Google's PPC, Gambling sites etc. I'd love to see how they manage things, and if we take data capture aside I suspect what they've built in house with the database vendors are in many ways far beyond anything a "web analytics" vendor offers.

Visual Insight: No. Web analytics data visualisation is not that special, it really isn't. Think about medical applications, pharmaceutical applications, networking applications, hell, think about excel... Yes, online has to handle the volume, but exceptional in its insight? I don't think so.

Trending Insight: Time series - do online vendors even offer time series reporting? Can you do anything more than simply smooth a graph? Can you predict future sales, page views click through etc? Can you do what if scenarios and see what effect campaigns are likely to have on your web performance? I think not, and I think offline has this.

Customer Segmentation: Do online vendors organise the data in terms of customers, then split the customers strategically to enable more effective marketing strategies? Nope. They don't. This is partly technological - still prefer to work with hits, then pages, then visits, then visitors, then users and finally customers but also cultural, a customer database isn't usually plugged into the online system, and besides web vendors prefer to collect data real time, rather than incorporating old data for analysis. Customer segmentation should lead directly to product development and marketing strategy - I don't often see web analytics currently driving this kind of thing hard within the organisations I meet unlike the offline companies I've worked with in the past.

Speed and Flexibility: The web has speed. Things happen fast from a consumer perspective. Stores can change their format overnight, Direct marketing that would have take weeks can be performed within hours, negative feedback can be immediate. Your weeks pay can be deposited in your bank and spent by you before you even leave the office. Admittedly products still have to wait while their are delivered rather than carried out of shop, but on the whole the experience is much more immediate, not only that new channels of information and interaction are cropping up everywhere. This means that marketing and the subsequent reaction to customer has to happen all the more quickly. How you do this and analyse it are actually no different to offline, but the speed introduces a number of other issues, namely that an organisation needs to be more structured, more organised, more focused on its USP and customer segments in order to provide clear messaging. The web has also always led in terms of automating solutions based on analytics when the data is available. (Collaborative filtering was perfect example of this, the likes of NetPerceptions leading before they dropped by the wayside). This is the key area where online is getting ahead of its offline partners.

Analytics: Offline have been producing predictive models that lead to insight, score potential customers, segment a customer base for years. Traditional, models and insight have been the domain of the analytics team. "Offline" software vendors are now realising that their applications must self learn or at least self maintain in order to keep the best figures in the systems. These models may combine propensity scores, risk mitigation strategies, product recommendations, contact strategies across multiple channels, tactical actions etc. Is online really any more complicated? No, I don't think - it has made the leap to automation quicker (think Omniture 1:1) but that doesn't always mean much as it cuts out the learnings to be had from a step by step analytics process. Oh and another thing, direct marketing was doing "MVT" 20 years ago, albeit with longer test cycles.

Market Research: Is tracking someone online as good as sitting down with them and asking to perform an action in front of you. I'm sure this has been documented elsewhere. On the web, if you can get enough users you can observe a multitude of behaviours and probably the ones you need to investigate. When analysing these behaviours online, do we really take the care and look at these things in a way more complicated that the better market researchers with their knowledge of sampling, customer segments, factor analysis? Again, the web has more data, but less quality of analysis, at the moment.

So on the whole, no I don't think that web analytics is better, I never have. Its strengths are 1) data collection 2) speed of interaction with customer 3) automation, but its lack of emphasis around the customer & product and the way these interact to produce coherent product and marketing strategies leads me to suggest there are still things to be learned.

One admirable thing is that the individual web vendors are trying to get good at all of the above, whereas I'm comparing them against the whole of business analytics vendors. That makes it a tough comparison as the hurdle to be climbed by the vendors is a lot higher, I just find it amazing that older lessons always seem to have to be re-learnt on the web.

As I hope above has shown, I think I can afford to keep blogging about process, strategy and analytics based on my years before Omniture and Adobe and still remain relevant without fearing that I will be stepping on any new intellectual property - I'll keep away from company secrets, I promise Mr HR person :-)

Internet Loyalty Cards

Where are they?

I often look for parallels in ecommerce with bricks-n-mortar stores and I've been wondering for a while, where are the internet loyalty cards? cards that would allow you to get benefits by spending at combinations of retailers?

Of course you have the credit cards, the amex's with their airmiles, but I'm more thinking about the likes of Nectar.

I'm pretty sure around the time of the first boom (2000) there were a few options being touted but since then, its never happened. Is this becuase they were never any use to businesses? I dont believe that. Is it because the technology is lacking? I'm not sure. I suppose one thing with a loyalty card is that it cant be replicated or shared as easily as it could on the net, but I expect this could be overcome with various cookies for the majority of customers.

I think the broader segmentation and multi-vendor offers could be of as much use to businesses online as offline, but I just dont think most companies are thinking that way as they dont see online users as "customers" but as "visitors that need to be converted".

Dont know really, maybe I'll come back to this thought later.

New / Returning Visitors metric

Its a good metric / segmentation.

Its not a great metric / segmentation.

Its got everything to do with the webchannel and less to do with your customers (if you're multi-channel)

Is a new visitor

1. someone who's found you via Search but never heard of you before and never been on your site.
2. someone who's used your brand term to find you but has never been on this particular site of yours.
3. someone whose purchase in your bricks-and-mortar shops religiously for the past ten years but who's only just found out how to use the internet.
4. someone who clears their cookies because they are wanted by the FBI.
   

Lets not forget, the technology has its limits - dont be guided by it and be prepared to ask the right questions. For the record I've always be willing to use qualitative data (surveys) on a sample of new customers to find out in particular what proportions I have in my technically "new" segment.

Homepage Conversion and Classic Consulting Recommendations

How often have been in a presentation when a web analytics “guru” has shown you your own homepage and pitched the following scenario:

“You have two banners on your homepage. Banner 1 converts (sells) much more of product A than Banner 2 converts for Product B. However, you get more profit per sale from Product B. If you swap the two banners around and conversion remains the same as the original banner (and why wouldn’t it??) we predict you could make $$$!”


This is what I describe as a “show them the money” pitch. On the surface, it works. Moving things around could produce more money and whilst doing it, you would be seen to be taking action. The guru has done their job and provided “insight” off the back of a couple of conversion metrics.

This could actually be true, moving that banner might be the answer to all your ills, but let’s take a look at a number of things you should consider before making that switch permanent.

1. Test it. Testing is so often the best approach. This will allow you to mitigate the risks of a complete swap whilst gaining all the insight
2. Consider how popular the 2 products are in your other channels. Is product B really likely to sell as well as product A just because it’s in a different slot?
3. Consider any seasonal reasons why product B should not be given top slot.
4. If using Site Catalyst, consider using an ASI slot to understand how frequently those products are purchased online when not influenced by these banners on the homepage. If B beats A hands down, your guru may be onto something!
5. Consider your marketing plan – if most of your impressions are for marketing related to product A, moving that down the page could present substantial risk.
6. Consider the cost of online marketing in the “profit” calculations. Don’t just think CPA (Cost Per Acquisition), think profit or margin per online sale as often these online costs aren’t directly factored in.

So what am I saying? You should get bogged down in a ton of analysis in order to go with what sounded like a perfectly good idea?
No, what I’m saying is don’t let the “guru’s” off the hook by swallowing their first suggestion that may be more based on general opinion than real statistics about your specific website and business.
Feel free to suggest more things in the comment section to consider in this basic scenario!

My First Assignment, Suggesting Action from Insight

I was fresh out of university in 1999 from a Cognitive Science degree and had my first assignment measuring web traffic for an ISP (Internet Service Provider).
I’d got about 3 days to come up with something interesting that could be used by the business to promote web analytics.

I came up with interesting stats about who did what, where and how they got to the site, but found what I thought would be the killer stat.

The ISP had a 4 page registration process and I’d found that about 50% of the time a customer reloaded the 2 stage of the process, not only that, some reloaded 3 or even 4 times with many with many people then dropping out at this stage. I looked at this stage closely, there wasn’t much to it – it was all about selecting your logon details and some minor personal info. I went through the process myself a few times and then suddenly I had my first work related eureka moment. It was very likely that these people are having the page reload once the usernames they originally select are thrown out because someone else has already picked them.

I write up my findings and present this to my business contact indicating that the registration Key Performance Indicator (KPI) could be improved, suggesting they provide customers with recommendations of alternative names they may like once the first try fails. (common now, but not something sites did in 1999!).

My business contact looked at it, considered it and took it away for conversation with his colleagues. A week later I spoke to him asking whether they would be making the changes to the process based on my recommendations.

He responded to me, the young green graduate suggesting that yes, the change would come but not for a number of weeks. When I asked him why it would take so long he said that thanks to my discovery his small team would be spending their time convincing advertisers who place their adverts on these pages to spend more money, because they could guarantee that for 50% of the people who see this page, they would see their advert more than once in a visit – the “fix” would come later...
Analysts often find it hard to suggest actions from their KPI’s and analysis. This isn’t due to those figures not being actionable, in many cases I’ve seen that analysts make good suggestions – just not ones that respond to the needs of the business they serve, and if they aren’t great, people dont listen.

If you’re a web analyst struggling with providing actions from your KPIs and insight, get to know more about the business requirements and strategies of departments that your analytics serves – from there your suggestions will become more relevant and more regularly adopted.

The cycle of analytics [DRAFT]

The Stages of Analytics

1) Guys, we need to report X - give me a trended chart that allows me to compare it over different periods

2) Guys, that was good but what I really need is to understand how to influence the trend, tell me more about it by breaking the trend over serveral relevant dimensions

3) Guys, that still isnt telling me why things are happening, what are you going to do about it? Oh build a multidemensional database? sounds good, here's some cash.

4) Guys, did I not say I wanted these results asap I want to understand 'why'? Where are you with this database thing? Oh, you found some more data and you want to create a "warehouse"?

5) Guys, I really want to start seeing something from my own data, just tell how to increase X. What do you mean too many dimensions? You what? You want to spend money on a fancy front end to display the data looking for insight? jeez, just get me something please!!

6) Ok, I like the sound of this statistical stuff, it sounds very clever. What does it do again? Prevent you from having to manually look at all the different dimensions in the fancy interface I bought you and what? finally shows me how to improve my chart?

7) Right, so this stats thing, its modelling, scoring and I "think" improving my trend, but I'm not sure as it all seems a bit complex and I dont hear anything from you any more. Can't it just tell me something simple about my business so I can do some proper actions myself?

8) What do you mean no one else understands it either?

9) Guys, you've really pissed me off. I've got loads of kit, loads of data but no answers, all I really need is to understand how to influence the trend. I tell you what, I'm going to get an "Industry Expert" in to do some brainstorming and build some decent segment. They might even do some surveying of my customer base instead. That seems simpler...

Best bloggers I've seen recently

I really like the Ad Contrarian and the Grumpy Brit.

Two very strong opinions that make for good reading. There's a general cynicism that I latch onto but on the whole, they talk a lot of sense - also, they dont get bogged down in stupid arguments about "analytics" more about what customers are trying to do and how analytics can be used to back them up!

Check them out!

Ad Contrarian and the Grumpy Brit!

and my favourite video (came via the Ad Contrarian).

Adobe + Omniture

Yeah, I've been quiet for a while.

I then thought better of posting a knee jerk reaction to the purchase by adobe. Although I normally have a talent for saying the wrong thing at the wrong time I actually impressed myself with my restraint and hopefully kept my job because of it.

Needless to say I do have an opinion on the purchase, I think its a good thing but perhaps not for all of the blue sky thinking that seem to bounce around the net whilst people tried to make sense of two companies with relatively different USPs and business models coming together. (btw. adam greco's was my favourite)

I'll I'm going to say is that when one company buys another, its the buyer that sets the agenda. I think there will be a time of reflection whereby Omniture will need to prove its potential growth, expansion and revenue to its new owners before it can set a new future.

I look forward to becoming an Adobe employee!

An Interesting Week in Predictive Analytics / Data Mining

Well, the big news splashed all over the front pages of the industry is that IBM purchased SPSS for $1.2BN. This is a huge deal for the industry.

Consolidation had been occurring in the more BI focused market with Hyperion, Business Objects and Cognos all being bought in the recent period but moves into the "Predictive Analytics" space was yet to happen. SPSS have in recent years made repeated announcements about their closeness to IBM, but I think the giveaway for some sort of takeover was the re branding of all products so they no longer used the SPSS moniker (Everything changed to PASW ...). Still, I'm surprised at the timings (credit crunch) and the price (42% above share price)!

This article in Computer World shows some of the more interesting stats about the deal.

For SPSS, it means many things but I'm always more curious about the shape of things to come for clever technologies in the marketplace and also things from an application or technical point of view rather than from a story of the companies finances. I realise that SPSS can now start tying its tools and applications very closely to specific technology vendor - hopefully developing bigger, better and more specific applications, allowing it to put predictive analytics into more useful (and no doubt more expensive) situations. I will be curious if this comes at a cost of its open attitude when working with other technologies and databases which I've always considered one of its strong points.

I'm very curious to see what happens and I'd also like to congratulate Jack Noonan for not only becoming a very rich man but also for having sold his company over the $1BN mark which has for so long been his goal.

And in other news...


The Netflix contest has ended - nearly... There's an interesting review of the ending on KDD. The winners still have to be chosen and I do wonder how their performances will fair in the final stages. Obviously the secret Test set has to be similar to the Quiz set but you do have to wonder which bits of the magic 10% that has been achieved by the winners occur due to overfitting/simiarities in the public data that are not to be found in the Test data. I've seen many models and situations where there inability to generalise has left me with a quandary of having to dump my "favoured" model.

In this case, how will Netflix decide who to pay the money to and will this have any interesting discoveries for the community for how we should consider validating our progress. If the current "winners" turn out to be well down the league in terms of test performance, and someone else steals a march on them, it could make for interesting conversations given the effort these teams have no doubt put it.

Who knows, but I'm all ears!

Data Quality and the Web

Avinash publishes another blog article regarding data quality.

Its a good article, nicely written but I have to agree with Mickey

"This is a nice sentiment but nothing new in the BI area - another example of web analytics not learning from “old school” analytics.

Although operational databases are usually very accurate, of the 100’s of customer databases I’ve data-mined over the past ten years, I cant think of many that were over 95% accurate, with customers data being ommitted, duplicated, mis-inputted etc.

The sentiment is right, go for precision in the data that helps you make your decision"

I get rather frustrated with web analytics going over ground that we have previously covered. Aviniash claims to come from the BI world. Unless he was working off an operational dataset, dealing with problems of accuracy should have been his bread and butter and therefore this should be nothing new for him, or for anyone else in the industry. If it is, I'll again quote Mickey,

"but will this be a cultural change for most marketers? It shouldn’t be. If it is news for you, either your tech people have been lying to you about you data quality for years or you havent been collecting much data about the customer"

Linking to a new blog

Just a quick post linking to a blogger who's rapidly becoming one of my favs in this area, the Ad Contrarian.

And here's a good recent article

Enjoy!

Tracking twitter and using other data to enhance your web analytics

In recent posts 1, 2 & 3 2 of my colleagues have gone into great detail about tracking twitter.

The response to these posts and presentation of this material at our Omniture summit was apparently overwhelming - everyone wanted to know about how they could track and use twitter.

I felt these posts were fantastic at conveying how one could go about pulling this twitter data into omniture and making use of it. What occurred to me was that peoples interest was not some much about the fact that it could be pulled into SiteCatalyst, but that the data could be got at at all!

It strikes me that people didn't realise that APIs could be used to get this kind of data - hopefully Adam and Ben have shown getting the data is not only possible but a very good opportunity.

Once the data is in Omniture, it's not tied to any other behavioural activity, it really is just a trend of data over time for a given segment, keyword or set of keywords. This could just as easily be pulled out of the API and into your own database as it would be uploaded to Site Catalyst.

In order to "analyse" this twitter data you would want to compare trends side by side and see if you can determine any kind of causality between you online activities and the various "tweets" about your company (or vice versa).

Personally, I would love to do a proof-of-concept to run a "text mining" tool over the tweets and see whether they could be segmented in someway to provide more meaningful trends than just a general indication of volume. I think the text mining software may have some difficulty with the online vocabulary (lexicon) that might need to be used along with the varieties of grammar that we see online, but it would certainly be interesting.

However, my main point is that many different data sources are online now that can be incorporated into analysing the customer. Because other tools make that data available you should be looking to take advantage of that somehow. Over the coming weeks I will be attempting to pull together a list of the online data sources I would consider to be most useful to analysing customer behaviour or at least comparing trends.

Remember, unless you can map the data from these tools directly onto your own customers you are always going to be speculating about their causality. Once I get this "ideal" list of data sources, I'll then look at ways of joining the data in that list to your customer data.

Drive your web development by responding to customer needs

A recent blog by Shane Atchison highlighted the issues in measuring social media.

One of the main things I'd like to highlight is right at the end.

The real challenges aren't technological, but operational. You have to define the right governance policies to manage customer engagement through social media, and you have to build the right workflow to prioritize responses, route information to the right people, and manage your content even as you release it into the wild.

I think this is the biggest and most important issue. When this isnt addressed it leads to a disjointed set of short term activities sometimes working against each other - the question is how do you join all the activities together in a way that makes sense?

For me they are joined together by the customer and their wants and needs. I find that if you can break down customers into a variety of key segments that revolve your activity around what they want and at what stage they are in their lifecycle with you then it becomes very easy to work out a series of practical approaches to dealing with these customers. You can then work out a marketing strategy for each segment and execute on it measuring the success of this activity.

You then measure the success of the activity you perform (site,ppc,twitter,facebook etc.) by asking

a) are you serving your customers needs in the best way possible
b) are you attracting enough volumes of the kind of customers you want

I find having segments to focus on can lead to a much better coordinated plan of attack, analysis and governance for you customers rather than flailing around with generic KPIs that most of the time are just telling you about how often content is being used.

Very few online analytics people focus on key customer segments however this for me is the best way fowards. Do yourself a favour and work out your key segments today, once they are in place most other things should fall into place.

Recommendation Engines, again...

It's been a while since my last post so lets get some simple things out the way.

Recommendation engines are increasingly becoming big news in the web arena. I've commented about this before but I'll repeat a few things if you missed them.

There is no sure fire single strategy to creating the best recommending engine, different products/content/activities in different markets are simply sold in different ways. However, in the mass retail market there is definetly huge value that can be added by adding those extra products to the basket and a lot of this can be done through automation.

So next time a software vendor tries to sell you an "engine" ask yourself whether it can be configured to adapt to the following 4 areas that you need to build a good recommendation engine:-


Business Rules: product placement through business strategy and good decision management
Product affinities: products purchase or seen with other product, similar price points etc.
Customer Segmentation 1: what do customers who look like X generally want?
Customer Segmentation 2: what do customers who look like X want right now based on all the recent behavioural information we have?


There are others as previously mentioned, but I think you need to be considering how these areas are catered for, analysed and optimised to ensure you get your greatest ROI for your software purchase.

Carry on recommending!