Last December I wrote a post on shopping process. One of the main premises on that post was that e-commerce should be entertaining, it should somehow captivate the customers and drive them through the entire thing while making it an enjoyable experience:
Buying stuff should be fun. Or as close to fun as it can get. It must be a pleasurable experience. Once we understand that our abandonments will invariably go down.
So I click the link and head over to this newly-found reading material. What do I see? The very first post I set my eyes on is talking about e-commerce, and from a perspective that has several contact points with what my personal opinion is.
(...) during my round trip to and from the milk, I munch a free sample, and grab a box of fudgesicles, knowing that my kids’ enthusiasm will counteract my wife’s annoyance.
When we build e-commerce sites, certainly we need to know what the customer wants, and give it to them. But we also need to use our spiffiest analytical tools to optimize two things: profit (dollars, not percent) and Net Promoter Score.
So, what would happen if we combine this thoughts on making the Shopping process entertaining and tempting our customers with extra treats with that other thing that has become a buzzword lately: Targeted ads.
Forget about the “ads” part for the time being. Lets just concentrate just on the Targeted part instead.
Using smart analytics, a powerful CMS and some multivariate wizardry e-commerce sites have the potential to create a unique experience for each customer that arrives to the site.
E-commerce has three main advantages over Grocery stores: 1) you don’t have to physically move products from one side to another, shifting aisles and pushing fridges, 2) you can track every single visitor and see how they behave and 3) you know where the visitors come from and, to a certain extent, in some cases even why.
With that in mind it is quite natural to imagine scenarios where visitor segmentation serves the purpose to profile each visit and build the e-commerce experience accordingly.
Granted, the approach would require massive investment in both Analytics and CMS, but the payoff should be huge. If the analytics team can profile and breakup visitors into smartly differentiated groups (natural search visitors, ad visitors, affiliate program referrals, coupon page referrals, etc.), pass along that information to the publishing / developer / user experience people and they, in turn can create experience that present stuff in ways that maximize the buying potential of each segment you’d have a winning recipe.
Multivariate tests should help determine what works for each segment.
In-site behavior should also be tracked, studied and used to present the visitors with different options according to the path they take.
This can be taken to various levels of complexity, and an investment & experimentation to revenue ratio would be determined. In other words: how much to invest, experiment and segment to obtain the highest return.
Another ingredient comes from a suggestion Mark made in the comments of my December post:
Reading what people write about your site shopping experience can fill in the gaps in analytics. Sites like Bizrate gather customer comments, and of course, blogs and forums are another great place to learn about the barriers to purchase one may have unknowingly created.
So Social Media (monitoring) can also play an active role in enhancing conversion rates and customer spending on e-commerce websites.
The final piece would be a dashboard that "adjusts" the settings under special circumstances. It is not the same to have an e-commerce site during the seasons than in Mid September when nothing happens. Since such events can be planned, they should be planned.
Stir together, cook for 90 minutes and you’ll end up with a “Fluid e-commerce experience”.
We don’t all have the same tastes, why should our shopping experience be dull and unique for all the population? The tools exist, the expertise exists, and some sites already run similar experiments.
Finally I encourage you to go ahead to his site and subscribe to the feed. It looks like we have another very eclectic blog at hand.
The event was scheduled at 9:30 AM, and it only started at 10:30... oh well.
"I want to share with you what is possible to do through the web"
Avinash starts by presenting his book, "Web Analytics an Hour a Day". People pay for something that is free (through his blog). He started as a blog. He is passionate about it, takes risks.
"It is astounding that a company (google) that does no Advertising has so much power". Well, the "no advertising" is relative, IMHO, we see google everywhere, on every online add, ain't that advertising of sorts?
"Faith Based Initiatives": How can the success of an Ad in a magazine be measured? An ad on Yahoo's homepage for the same product is not a faith based initiative, because it is relevant and can be measured.
Online Marketing: not faith based initiative.
"I am not telling you not to run ads on magazines, just telling you that it is not as accountable as the web".
As with everything in life analytics can be as complicated as you want, but it is easy to start with it.
"Google holds your trust with us as the most important thing we can ever have, we realize how fragile it is. If you share your Analytics data is just used to benchmark; if you choose not to share there is a legally binding term that prevents us to share it".
"Nothing I'm going to share with you is unique to Google Analytics, you can use any tool"
Showing Geo segmentation, content segmentation, how to break down data within Google Analytics.
"Bounce rate is one of my favorite metrics. I think it is a sexy metric, I love it because beyond pageviews and visits (which are analytics currency) it shows the user experience, it shows how much you suck"
"I came, i puked, i left, that is bounce rate"
"It is also a very actionable item."
"Look at the top entry pages in your website and see which of those suck".
Avinash is analyzing Clarin.com; it has 419 links, not counting buttons. They fotocopy the newspaper and make it a website. Shares http://tr.im/hmSG which is an article that tells where newspapers are going.
Reports should be focused on outcomes. What you should care about (and what your boss cares about) is how much money you made.
"We improved customers satisfaction by n%" is another possible outcome, that can be shared.
"Life is not like a one night stand, people come many times to your website. You have to understand the behavior of the customers on your website" (picture of a conversion funnel).
"3 pages can account for most bounces. You can fire most of your company, fix those 3 pages and make a lot more conversions"
"It is idiotic to measure the number of members, it is not important how many sign up to your site (Using facebook as example, but when they start actioning on the site. If you are Facebook you want people to come over and over again, you need to analyze whether people spend a lot of time on the site or not"
Shows an example that shows that 66% of people return within the day to the website"
"A newspaper should measure the depth of the visits, this makes you think very differently about your business"
"See what the important segment of people need and want from your site. Take a look at people who puked and left, try to understand why they left"
"70% of the people working on a Newspaper we recently analyzed were writing content with a high bounce and no - return rate"
"The average convertion rate for the US on e-commerce sites is 1.72%. What was happening to the other 98%? It is very important that you quantify the value of that 98%. You can set different goals to see if there are other types of conversions"
"In my blog I measure hits to my about page, it makes me happy, in your company measure the ammount of people who visit the executives page, they will love it, they wont admit it, but they'll love it."
"All of the clicks on my blog are tagged, because the conversions happen elsewhere"
"The web is also reducing costs, there are a lot of different conversions of different types going on"
"Look at the whole company, always compute the total economic value of the site. For example, in my last company we saved $3,000 in consultants each time someone applied jor a job through the website"
"If you do adwords don't just measure visits and pageviews"
"Getting Visits is OK, making money is astounding"
"In my blog I had 64K visitors, 23K came from Search Engines, searching 11,5K different keywords. That is the power of the web. How do you find those keywords? You can use 'search based keyword tool', what the tool does is bringing the search queries and the indexed pages back together"
"Custom reporting: every individual needs and is interesting in a different set of metrics"
"When you don't segment your data don't be surprised when you can't make any important decisions"
Last part of the the talk: "Fail Faster; Experiment or 'Die'"
"Most web sites suck because the HIPPOS (Highest Paid Person's Opinion) created them. Never let your Hippo think they are the customer of the website. You need to prove them wrong fast. I recommend you use google website optimizer. I worked with a company that reduced the number of ads in their web page by 30% and increased their ROI"
"Don't guess, don't impose, Partner."
Analyze trends, Identify Audience in smarter ways. You can even look at the competitor's trends and audience of your competition.
"If you share your data on Google Analytics you get access to 6 different metrics as benchmarks"
"Privacy is extremely important. People will share data with you if you are clear enough about what you are collecting. Google has created a plugin that prevents google from collecting data. We believe that Yahoo! and Microsoft will follow this trust mechanisms."
"A+B testing makes each website more unique. A good example in teh US is if you go to BestBuy and CircuitCity websites, they look alike. Circuit City filled for bankrupcy, Best Buy invests on targeting their audience and testing to optimize the site for their audience"
"There are few things that are shared by most sites. Internal searh (on top right) is one of them. Carts are usually to the right, but a site improved conversions by moving it to the left"
Web Marketing shouldn't be a part of marketing at all. Why so? Because Web Marketing is measurable, those holding the threads and running campaigns and initiatives can be held accountable for the failures and successes they encounter. Traditional marketing works more like a religion, companies grant faith (dollars) and hope to get something in exchange in the afterlife. That usually yields the eternal discussion of wether bucks spent on Marketing are investment or just expense.
Web Marketing, on the other hand, has a full range of tools and methods that can measure the ROI to a very acceptable degree of accuracy. So, with the focus put on dollars, measurement on web builds us the case, and, with some good fortune, gets more funds to invest.
This has become known territory for all of us working at web marketing. We spend a lot of time, effort and money to get the greatest and latest in Analytics tools because we have learnt that it pays off.
But I do get a feeling that haunts me. The knowledge, deep inside, that we have just scratched the surface; that there is a ton more that we can learn with the different metrics we can acquire, a lot of things that are more complex than cash-in and cash-out, a knowledge that can -potentially- lead to a better understanding of the web audience and, eventually, profits. (Gosh, is that phrase messed up or what?).
If you read parts 1 and 2 of this series you probably know where I'm heading. If you haven't, go back and read!!
We sin. We do the exact opposite thing to what I have suggested on this series. When looking at our facts and figures, those thousands or million visitors are just sheer numbers, abstract entities that build up the charts. They are "visitors" who make "visits" and, if planets are aligned appropriately, they generate a "conversion", else we have an "abandonment". A -> B -> C or D. Analysis can go deeper, try to determine what on the face of the earth influenced the conversion or abandonment and where and how it happened.
With some fortune, some multivariate tests can take place as well. Those determine if a rather limited set of variables building some recipes are more successful in terms of driving our visitors towards our goals.
I know I'm making this sound simplistic. Getting this much takes a lot of time and hard work. Code has to be developed, metrics tools need to be configured, campaigns and start-ups need to be tracked and measured and analysts need to make some sense out of it all.
So, what we have done so far is trying to figure out how things work from the company's point of view, but what would happen if we tried to understand the other end of the function?
What we should try to gather data from and analyze is the public "sentiment" of people reaching out to our web-based efforts. Social Media Marketing does this already, to some extent. Buzz reports, blog and forum monitoring and similar initiatives all provide insights into what's going on on the public perception end. Yet, for the most part, the public is still treated as a mass.
It is not viable to treat each of a million different visitors as a unique entity. For starters the sheer numbers are overwhelming. Then again we are not able to pick up enough data to get the complete picture of why every single individual does what it does. Thus, analyzing each individual is an utopia.
So now we stand in the middle of two opposite approaches. One too general, too broad, too focused on businesses' own belly buttons. The other an utopia, an unviable approach.
In an almost Buddhist reasoning, the truth lies somewhere in between. Focusing metrics on the users, taking advantage of the vast amounts of data we have at hand, learning how to serve them best in order to achieve better business results.
Easier said than done, since we're usually too immersed into trying to close a good quarter to stick our heads out of our own narrow vision to try to establish policies that will, most certainly, not yield results for a rather long period of time (in business terms, that is).
More on this line of thought will come, for certain. For the time being this series is done.
I am always amazed by how ideas can come from anywhere. That is one of the reasons why me reading interests can span anything from Astronomy to Linguistics.
One of Chris Locke's blogs, Mystic Bourgeoisie, had a reference to a BBC documentary called "The Trap". The 3 hour piece is far too interesting and far too long to make any good summaries in here. That being said, I shall share the concept that draw me to write this lines.
During the Cold War, mathematicians developed model of society based on game theory. For this model to work, the human beings are reduced to a simplified version of ourselves. This was latter applied as the theoretical basis for multiple government policies put in place during the 90's. I really recommend the documentary, wether you agree or disagree with the content, it is most certainly quite eye-opening.
Back in the corporate world an analogy can be drawn as well. For decades companies have regarded customers as little more than potentially-buying machines that can be manipulated into acting in the desired way by the means of advertising and marketing. The advent of Internet and social media have built greater complexity into this model. The relation between companies and customers is far more complex and far more difficult than it used to be.
For starters, the one-way communication model of traditional advertising seems to be sinking rapidly into uselessness. The never ending bombardment with TV, Radio and Print ads, billboards and, dare I say it, on-line advertising has made the mean population almost numb to such means of influence. There have most certainly attempts to insufflate some life back into traditional ads, viral being the latest. The funny thing is that Viral has to rely on the user's action to be successful, something that directly contradicts the old paradigm of "we talk, you listen (and buy)". Yet "viral" is still a traditional way of promotion in most senses.
While I was giving some thought to this concepts I couldn't avoid but thinking about Isaac Asimov's "Foundation" series. In case you haven't read it, I'd say you skip the next paragraph, I'm going to spoil the story.
Asimov thinks about a future galactic empire that is swirling down in decadence. No one seems to notice that, but a group of thinkers, commanded by Hari Seldon, creator of the Psychohistory field of thought. Psychohistorians build mathematical models that can predict the future, thus they foresee the decadence and collapse of the empire and, more importantly, the dark ages that are to follow it. Their plan to reduce this dark times is to build Foundation, an encyclopedic world, a stronghold of knowledge, guided through the ages by the sporadic recorded messages of Hari Seldon, that appear every time a crisis is faced by foundation. All goes well until something that no model can predict happens: a mutant -the Mule- has the power to manipulate human brains at will. He overtakes over foundation and puts Psychohistorians plans at stake. A Second Foundation exists to deal with such issues, their field of work is not maths, but developing human's mental capabilities.
The shifts in paradigms are not an easy thing to cope with, the belief systems associated with them are usually deeply rooted within people advocating for them. change is not easy.
Part 2 of this series will deal with the social media aspect that surfaces from this line of thought.
Either if you are a private blogger or if you have anything to do with corporate blogs you'll crave on increasing traffic. Whether it is for pure ego or for a business need, having people to actually come and read your blog is a good thing.
If you talk to any traditional on-line Marketer they'll say that SEO is the way to get such traffic. And it is true, to some extent. I won't write a SEO article giving out tips and tricks on optimizing a blog or site, others have done that before and in a far better way than I could try to achieve. If you are interested in getting some SEO knowledge I'd recommend you reading and looking around at Aaron Wall's SEO book.
The WordPress community has built several plugins that help optimizing posts and sites with crawlers and robots in mind. You can add keywords, meta tags, sitemaps and a myriad other things to make things more understandable for search engine robots. Yet things are always evolving in the SEO world and new tricks surface almost on a daily basis.
But, as Dylan would say the times they are a changin’. Blogs, wikis and other rich or social media sites don’t rely that much in SEO to attract traffic, since they are more socially engaged. People tend to find them on a slightly different way from the way we're used to find other things on the web (like a bargain or information on a particular subject, for example).
For blogs particularly the story is quite different. People tend to visit blogs not so much thanks to SEO (although it does play a role in a certain percentage of visits) but more thanks to links, comments, references and social finding & bookmarking sites (Slashdot, Digg, del.icio.us).
Here is an interesting note on how things are connected. As I was writing this entry, news came in from the Pub Owner on a related story. Here’s a short excerpt:
More important than that traffic, however, was the list's role as "linkbait." Users on Digg and other social media sites created more than 800 links to Olthuis' list in forums and blogs around the Web. Because Google ranks a Web site's relevance based on the number of other sites linking to it, LifeInsure now ranks fourth in Google's results when the search giant's millions of users search for "life insurance." Suddenly, the company had free advertising that put its name right next to huge brands like Metlife and Prudential.
You can read the whole article on forbes here.
(Sidenote: this really depicts how long some drafts take to come to life, Tim's mail came almost 2 months ago)
This proves that not only rules are quite different for this types of sites, but also that this rules, if played appropriately can also boost SEO. Of course it is no easy thing to get slashdotted or dugg (digged?), but that is a very different issue. One that I personally feel is more honest towards the end user (aka the reader), since it relies more on generating good content instead of trying to fool a robot and trying to make an engine believe you’re the best thing after sliced bread (after all, what do robots know, huh?).
So, the good news for bloggers is that they can get to spend more time delivering worthy content instead of trying to get google to put you on a higher pagerank. It most certainly will pay off better than just hacking a pagerank.
The thinking is certainly refreshing. I've grown an allergy towards sites that look more interested in catching search engines than catching me.
It is the tale of good content driving traffic, engaging readers and being the single best thing where to put focus on. If you are on a tight budget (or on no budget at all) and have to make choices, invest in the quality of what you are publishing.
As with all good things in life, the focus on content approach has some difficulties, particularly within companies. It is much harder to effectively measure. If you invest in paid keywords you can fairly easily crosslink the numbers you get, with good content measurement is not as direct and easy. It is also not immediate; some excellent blogs might take months (or even years) to take off. There is trust to be won. Then again you might get dugg on your second post ever and jump to fame almost immediately.
Please remind me to write a post on how to try to measure the ROI and KPI of good content. I have some wild thoughts that I need to thoroughly think, test and try out.
As I already said (quoted, actually) a couple of months ago, Content is the next killer application. As social bookmarking and other word-of-mouth tools evolve and become increasingly popular, and as search engines and indexation systems evolve to take more into account the real, human-aimed content the real SEO will focus more on delivering interesting, well thought and well put content and less on tricking search crawlers.
Bottom line is: trust your content, if you are genuine, honest and smart it’ll pay off much better than any other thing.
I don't think I've mentioned a lot of details on my new position. Half of it works around social media; I have talked a lot about it lately. What I haven't posted much about is metrics; the second half.
Metrics are a dangerous thing. I find them particularly appealing and obsessing myself. In the world of marketing it is hard to accurately measure anything, and that became the common rule. Yet the world of web marketing is quite different in that regard.
Analytics can potentially touch every single aspect of a web infrastructure, from design to revenue; from campaigns to usability. That potential is hardly ever fully used, though. The degree of integration with a CMS or other methods of publishing web content seem to be the most important roadblock one has to face. Luckily the trend lately seems to be that most content management systems have started contemplating the importance metrics have, therefore providing some toots to work things out.
Another hard thing is that there is no single approach towards data capture. Even within the same Analytics vendor there are quite some different routes that can be taken. Finding the right set of implementation policies is a high-demand activity; it requires quite some trial and error and a lot of tweaking the best practices to fit a certain market or implementation's need.
Even when one takes into account this and other difficulties, the way metrics pays off is unparalleled. I'm quite sure someone has made the math on what the ROI is for every buck spent on metrics & analytics for different industries, but I'm not aware of such studies mself and I'd really like to figure out that much on my own experience. I guess I'll add yet another thing to me to-do list.
I can easily picture a scenario where metrics are the glue keeping a web-site together and working. The risky thing is to fall in a metric –driven policy. Metrics tell us what happens and what the trends are; yet they should not be an objective per-se, just the way to understand how things develop.
There is a slight nomenclature difference that can make a huge difference in the focus and results. Although "we want to increase click through" and "we want to make valuable content easier to find" might yield the same result in terms of the raw numbers, the implementation methods would differ for both statements.
A site can easily increase click through rate through various methods. Eye-catching banners and phrases, reformulating content disposition, even could work. The click through percentage figure would increase at the end of the day; but you might end up doing the exact opposite thing to what would be advisable. No single metric and very few metric mashups should ever be a target on their own. Metrics are analysis tools, they reflect what happens and can, to a certain extent, foresee what's to come, but they should almost never be at the driver's seat, but rather the navigator's. Metrics and Analytics help to understand reality and help making smart decisions.
This is not an attack to metrics and analytics, quite on the contrary. Having such things in mind should preserve the integrity of the M&A department. You can't possibly understand what's going on and identify areas for improvement without metrics, but changes should not happen to increase or decrease a figure, but rather a figure should change because it is part of a bigger, comprehensive aim.
Aviansh Kaushik's blog presents "Ten Minutes With….. Mike Moran, IBM", an absolutely MUST for anyone with any level of involvement on Web.
One of my favourite phrases is Mike's reply to "How do you make yourself indispensable to a company?":
I think passion is a big part of it. Find what you are passionate about and go after it–you’ll do a better job. But you also need to find out what is important to your company.
So ABC... yet so overlooked.