Sentiment or attitude analysis is about understanding the mood of the person making a comment or post about a topic.

It’s been my opinion for some time that you can’t measure sentiment with any real confidence in the results, at least when you’re relying on the results of tools. Jim brilliantly illustrated this by showing some sentiment analytics from Twitrratr.

As you can see from the red strikeouts over 50% of the comments here could be categorized as falling into a different category.  Many of the negative comments about Obama here are actually positive but because they talk about “lose” or “won’t” (as in might lose or won’t win) they have been categorised as negative sentiment.

They clearly aren’t negative, they are actually positive “Tod sez Obama could still lose pa, ohio and fla and still get to 270, that’s a favourable playing field” is a positive slant saying Obama could win without those states. It was listed as negative. Twice!

The point here is that even humans can’t agree in many cases on what’s negative and what’s positive in terms of sentiment so how do we expect a machine to do it?

My advice is to treat “buzz” or “sentiment” analysis in a qualitative way. In other words manually analyze whether it’s good, bad or indifferent! Kwantic for instance developed its own in house tools to randomly select comments around key terms.

Our tools randomly select comments and then we use one of our market research professionals to score the context of this random sample. The tool collects enough comments to ensure statistical significance is there and our analysts do the rest.

To me this is the only way to analyze sentiment correctly because only a human can put it into context, get the language right, understand slang and give an accurate picture of what people are discussing.

Comments as always welcome!

Personally I find standing up in front of hundreds of people a nerve wracking experience every time I do speaking engagements so I can only imagine what it is like to put yourself in front of 19,000 people including the Finnish president Tarja Halonen. If she was nervous when she went out there it didn’t show as she got on with the job of battling and eventually defeating the Swedes.

I don’t know much about hockey truth be told but I do know about character, fight and spirit. Finns call it Sisu. Terhi and the Finnish team had plenty of that (as well as considerable skill) twice going ahead with superb conversions and twice being pegged back by a somewhat fortunate Sweden, at least that’s how I saw it.   When sudden death extra time ensued and the Finnish came out they looked hungrier, faster, fresher and far more dangerous than their Swedish counterparts. Suddenly it was all over. Victory was theirs and the celebrations began both on the ice and in the bars in Helsinki!

Salue! and enjoy the moment Terhi, they don’t come around too often and I know like anything else good in life it was hard earned and thoroughly deserved.

Let me first say I have no problem with that approach, especially in immature markets. It has served us well for the first 5 years of the WAA but I just feel in certain cases it is outgrowing this method. I don’t dispute the good work done to date by Jim, Aurelie, Rene, Miles, Dennis, Stephane and a ton of others I haven’t mentioned but do you really think that going forward a global organization based in the USA can service other countries as well as the US?

We had 300 people at the last WAA event in Finland and we got 2 sign-ups for the €1000 or so that was spent by the WAA on the event according to Petri - Finlands country manager.

Is that a good ROI for the WAA? I would argue it’s a very poor return on both the Finnish WAA funds for the year 2009 and for the global WAA. A CPA of €500 per sign up. So where is the mechanism to activate that crowd? How do we get them involved with the local community beyond free events? Where is the benefit for the WAA?

In Finland at least the groundwork of awareness has been generated (300 people is a good size), that’s the first job of a country group or country manager but when you get to a certain size you need more local organization to activate things and that costs real money to take it to the next level. That is the brutal reality.

A global organization can work if it is prepared to heavily invest in local operations. Usually global giants follow the model where the most successful countries are the ones where the steady funds are directed with smaller “exploratory” funds directed to high potential markets.

The WAA can’t do the same thing as it’s a non-profit, they simply don’t have the resources. There are only 2 ways to generate finance.

1) The WAA as it stands now funds operations. The question is how do you allocate budget? Kalle Heinonen submitted a budget request based on reality in 2008 to really build the Finnish membership and it was refused on the grounds that what we’d asked for was simply too high.

2) We generate local funding when certain levels of maturity are reached in different markets. Finland, Sweden, Denmark, France the UK and Spain could be focusing their efforts on supporting local memberships while paying a percentage of subscription fees back to the WAA in order to use the brand and have access to global standards, materials etc.

I am not trying to be controversial for the sake of it, all I am saying is that when a certain level of maturity is reached in the local market you need an increased level of professionalism and finance to run things. I am not saying an EUWAA would help either. I think it needs to be a kind of mature market by mature market kind of franchise.

1. Traffic generation via link farms, 1000s of domains, or “our network of websites”.
   When I heard this one I couldn’t believe it was still a tactic some agencies we’re employing.
   Good practice
   If you’re buying traffic from a network of branded content websites it’s good practice. If an agency pitches working with a well known brand name it’s fine (Ilta Sanomat, MTV3, Suomi24, YLE for instance). Look at TNS metrix advertiser list to get an idea of sites which are high volume in Finland. When an agency says we’ll drive traffic by advertising in a targeted way with traffic coming from (for instance) sites on this list then ask you to do ads and banner creative to work with them they’re probably doing an honest days work.
   Bad Practice
   Spamming search engines. If an agency says “we have a network of thousands of websites that can rank you in Google for your keywords” or “lets develop a network of 10,000 domains that link back to your brand” or “Google works by the amount of inbound links you get around the keywords you use, let’s therefore create 10,000 websites linking to you for your keywords”. If an agency comes at you with any of these ideas just show them the door. They are digital marketing cowboys. Spammers. Scam artists. Conmen. Parasites! There are ways these guys use to trick the search engines but it will usually only work temporarily until Google hunts down the offending sites and bans them. So if you’re at all worried about losing the current levels of traffic you get from Google then don’t do this. I first heard of these shady methods in 2002 and Google is on a constant hunt for the people that employ these tactics. The big problem with the parasites is that nothing happens to them when you get caught, because it’s your brand that is taking advantage of their tactics. It’s you the advertiser trying to get more traffic that gets caught and gets banned because it’s your site taking advantage of the inbound links. The parasite spammer just jumps to the next project.
2. CPA/C (cost per acquisition/click) Daylight Robbery
   Again when I heard people were still doing this to unsuspecting advertisers I thought I had been sent back in time to 2003/4 when I first saw this bad practice happening. I thought that the reputable agencies had long ago realized that stealing money from their clients in this way would only get them a bad reputation. I was wrong. One of the biggest agencies in Sweden is doing this right now and they deserve everything they get from the client I talked too. Again I’ll start with what I consider is good CPA marketing;
   Good Practice
   When an agency says that they will drive traffic to your top keywords via search engines and shows you the performance all the while comparing this to organic traffic coming to you for free from search then again they are trying to do an honest days work. This way you can judge whether the click cost is justified for certain keywords and can see particularly which brand keywords you don’t need to bid on Google for.
   Bad Practice
   The agency focuses purely on the SEM paid clicks and charges accordingly. So I heard that the parasites in question were bidding on a CPA basis on brand keywords. The brand keywords were the best performing ones and the cheapest to buy because the client in question had excellent SEO for the brand terms and none of the competition bid on the terms. So they are buying the keywords for a few cents per click and charging the client €25 per sale from those clicks when the client would most likely have gotten the clicks for free anyway from organic listings. A huge margin for the parasite. My client referred to this as “almost a crime”. I agree. Flogging is too good for these toe rags. It’s the same with clicks and click packages. Be careful if you hear “we’ll sell 1000 clicks for €1000″ because in many cases click prices are less than 20 cents each (unless they’re highly competed on) meaning your parasite is earning 80 cents to the euro from your media budgets. An easy thing to check is click prices for your keywords. Google has a CPC (cost per click) estimator where you simply enter your keywords and filter for CPC. Do that for your top 5 keywords before you talk to the agency so you understand the mark-up they are adding. Of course you have to pay for their time but make sure you know what you’re paying for and you’re not being ripped off.
3. Social media gurus
   These guys are multiplying like rabbits. So far I’ve heard everything from “Do you know what Twitter is?” to “Post flame wars to drive up your web traffic”. It hurts my head to think about it.
   Good Practice
   The point of social platforms is to be a focal point of a community or to facilitate conversation. Engagement is key with communities and conversations. Social strategy should first take into account how you’re going to engage with the audience in a meaningful and useful way to that audience. Then measuring the impact of that engagement. It’s a very qualitative process and not like broadcasting your message like advertisers have done in the past. If an agency comes to you with a strategy that says you’re going to have to employ people full time to engage in long term social activities for your company then you have an honest company in front of you. Viral campaigns across social platforms can work as well but should only be part of longer term strategies that effect your brand, because it all affects your brand. Social should involve a lot of man hours because it’s about people communicating with people.
   Bad Practice
   Set-up a Facebook page and watch the traffic roll in! Get a Twitter account and use robots to locate followers and force them to follow you. Automatically get 1 million friends & followers, social automation. If you hear anything about social media that involves robots be careful. Search engine marketing parasites like the ones I mentioned in the above examples have also jumped on ways to spam social marketing as well. The affect to your brand could be disastrous if you ignore the bad word of mouth that can quickly be generated. People quickly complain. I for instance am using social media right now to engage with you to tell you I am pissed off with the digital marketing parasites that are plaguing the industry. That could happen to your brand and once again it won’t be the “social media guru” who deployed the spam bot, it will be your brand that gets the negative effect. Parasites usually survive in life and in digital marketing unfortunately.

The parasites give the digital marketing industry a bad name. Comments? Opinions? Rants?

If I could only do 5 things using Analytics for my business in 2010 these would be the first things I’d do.

1. Measure segmented outcomes. In e-commerce it’s easy, measure conversion rates to sale across your different traffic sources. With lead generation measure conversion rates of how many people sign up to hear more about your services. With the media business the outcome might be the amount of heavily engaged visitors. Those that stay on your sites longest and see the most pages are able to see more adverts. By segmenting you’re able to determine the traffic sources to concentrate on (the one that drives the most outcomes).
2. Do statistical analysis. By doing statistical analysis you can pinpoint where you should look. It shows you where the “abnormal” trends are. A couple of years ago I wrote about a method to do this so rather than repeat it here, please check it out. Google Analytics also does this for you now (their intelligence feature) so if you use GA make sure you start there.
3. Measure segmented bounce/exit/abandonment rates. As a place to start monetizing your potential. These metrics indicate people who haven’t found what they wanted. In all three sectors you have different reasons for not wanting this to happen. In e-commerce sites you don’t close the sale if the abandon from your shopping cart process. In Lead generation sites you don’t get the lead if they exit from your form sign up page. In Media sites it’s not ideal from your revenue perspective as 1 page visits (especially) don’t add up to many advertising Euros. Segment here again to find the worst sources of traffic so you can see if it’s worth eliminating them or changing your content/processes to reflect what people want.
4. Include other data sources. You don’t live in a vacuum, your competitors might be the reason your conversion rate is low if you’re in e-commerce. What if they sell the same product you do at half your prices with twice your shipping speed? If you’re in the lead generation business check out whether you are outperforming your competition. You may be driving 20% more traffic (and thus getting 20% more leads) than last year which is a result but only if you outperform the rest of your market. If you’re in media advertising well you need external numbers to compare how well your competitors are driving traffic so you can put your own traffic into context.
5. Monetize, Monetize, Monetize. At Kwantic we monetize everything we can. In e-commerce it’s very easy, improving conversion = €xxx,xxx similarly with lead generation any improvement in exit/bounce/abandonment translates to potential value. With media sites you either cut your costs, increase your traffic or get more engaged traffic. Increased traffic or more engaged traffic means more advertising can be sold thus increasing revenue.

So, there you have it. Short and hopefully sweet. Only one thing left to say… Happy New Year!

I read the book cover to cover in the trips to and from the UK. This is an excellent read. I’ve read everything out there about analytics, I’ve written one book myself and found myself smiling throughout this read thinking… why didn’t I say that? Or, why couldn’t I have put it that way?

I am my own worst critic and my own natural tendency to compare my work to others had me feeling like my previous efforts with Cult Of Analytics were 7/10, must do better!

The sign of a good book is when you refer to it again and again after reading Web Analytics 2.0 that’s exactly what I found myself doing. As a consultant this book is a constant reminder to me why I got into Analytics in the first place.

SO… I’m not going to tell you what’s in it.

I’m not going to say that page 21 has some great questions to ask your analytics vendor (especially the last question). Nor will I refer to page 65 that has a good diagnosis of how to improve conversion, a subject very close to my heart. I’m not saying that Page 85 starts with the question I start all my KPI workshops with (What is the best source of traffic in terms of outcomes? – Ok that one I did tell you!).

The book also contains an excellent summary of ideas on how to gather and use competitive intelligence. But I’m not going to tell you what it says.

I’m not going to mention the content of page 160 about measuring economic value nor the other cultural tips Avinash smatters through the book and which are all applicable to any business.

I’m not even going to tell you his approach to measuring mobile phone usage, blogs, twitter and other so called Social Media tools. Excellent simply because a lot of this stuff is not written up yet in the simple style that Avinash uses to explain his ideas.

Finally I’m not going to tell you his strategies for measuring SEM or SEO or the way he combines the two strategies. Simple but very effective.

I’m not going to tell you because Avinash says it better. You should read it yourself and find out first hand.

Just the other day my friend Sean D’Souza put some material out there that used to be sold for $2500. I have advocated his materials for years and was impressed when I saw he was giving one of his old products away. I didn’t even think about it, I just tweeted the link to my followers.

This wasn’t something that Sean asked me to do, I just did it because I want to give people following me on Twitter something of value  and I know his product is good having been his customer in the past.

This is a way of selling that we’ve been doing since the stone age but now by utilizing social media companies can take advantage of social selling. We’ve always relied on first hand evidence from friends, colleagues, Mrs Jones or the caveman next door before making our own decisions on what to buy into.

Sean is asking you to buy into his idea, and removing any risk by giving you the product. He then hopes you’ll buy his book and his other range of products. In effect he is using old customers like me to tell you how good he is so that you’ll buy into his idea and then you make your own mind up.  I know this and I don’t mind.

With only a little planning, measurement and effort you can take what you sell and turn your customers into a free salesforce that is more powerful than any cold call, email, product pitch or slick sales talk.

With that in mind, do you have examples of social selling you would like to share? And are you optimizing them effectively?

Slightly bigger news from the other side of the world then!

There were 3 Omniture guys at the event and like me, none of them saw it coming or at first understood why it had happened. After reading reactions, particularly a good one here from Russ Mann and a very good one here from Eric Peterson as well as talking to Omniture I have reached my own conclusions.

1. This is a great deal for Omniture. Probably Josh just did the best deal of his career and you’ve got to applaud that. The Omniture management team basically managed to sell a company that is struggling under fierce competition from Google and Yahoo! free solutions as well as a number of lower priced good systems for a price that is 25% higher than its market value. $1.8BN is a superb price for a company yet to make a profit. This was my first reaction to hearing the news and that opinion hasn’t changed since.
2. Like Eric Peterson I don’t buy the synergy. Russ Mann makes some great points about how the technology “can be integrated” with Omniture. But what he fails to address is that while it might make it easier to code flash or flex it isn’t going to solve the fundamental cultural problems that exist in companies. Just because the designs have tags in them doesn’t mean that people will suddenly start using analytics. I have written 280 pages on this subject and could’ve written another 500! It’s a bigger problem than the technology.
3. Adobe have certainly taken a bold step and drawn a clear line in the sand to Microsoft and other competitors. It takes balls (the kind made of granite). I was surprised when Omniture bought Visual Sciences. That though was a completely understandable deal for Omniture and VS with easy to see wins on both sides. Adobe however are entering a market they really have very little idea about and the business models are totally different right now. The move by Adobe is certainly brave in my opinion. It remains to be seen whether it will be a foolish move.

The jury for me is out on this one.

I have my doubts that the market is ready for Adobiture.

No-one I know in the Analytics field uses Adobe products to any great extent. Adobe creative products are for the designers and coders and education in this sphere about Analytics is one of the big jobs facing the industry. If Adobe expect that to change then they will run into trouble in my opinion. Of course if this is a step in the direction of educating the creatives then that’s great!

I hope for Adobe’s sake they have done their due diligence on how they will work with analytics beyond their own product suites, actually in the enterprises using analytics. I can see Adobe products integrating with Omniture as both companies have the technical abilities to do that. However I think the bigger problem is educating the people using their products as well as those that don’t inside the enterprises that will be their customers.

If Adobe feel they need to revolutionize their products to include measurement, they have my support but I feel they could’ve done it a lot cheaper than acquiring the biggest vendor on the planet.

I have left Trainers’ House because the plans in 2009 did not match my ambitions. Their growth system story is good and I believe it can and will work. I would like to say good luck to those at Trainers’ House and thanks for 3 good years. I learned a lot about working with large enterprises, consulting, and had a great time there. I don’t have any hard feelings towards the company at all, I just felt that they were focusing their energy into a different direction than I need to go.

Different path

In July 2 serious opportunities were presented to me (among a number of ones I didn’t like the look of). One was to hold an executive position in a global company to help spearhead their analytics division which is already a large going concern and probably a big salary to match. I didn’t hesitate, I took the second option without wasting the time of the good folks who offered me the first. The second was to start again as a director with a new company. Those of you that know me, know that I have already been down the second path and may wonder about my eventual decision to do this again.

The reason is simple. I feel I have unfinished business. I have an itch to scratch. Can I help develop one of the best analysis and marketing companies in the world? Can I continue what I started in 2003 in the days of Aboavista? This remains to be seen. My targets however are to have 20 people working with me by the end of 2009 (with revenues to match). To double that by the end of 2010 and to continue to strive for excellence in all that we do.

Right now we don’t even have a logo, nor a website (well we do but it has no information on it). We do however have a concept. Quantic is a mathmatical term meaning “a rational, integral, homogeneous function of two or more variables.” (dictionary.com). In other words the rational result of 2 or more things. I’d like to think that our new company Kwantic will strive to get good results for our clients from doing analysis and taking action based on that analysis. Simple as that.

So please bear with me if this blog is a bit light on the content side for a while as I help build the foundations of the new venture and as always feel free to add your comments. Onward and upward!