How Agile can Analytics be?

As many practitioners and Project Managers know, managing SDLC projects has its own nuances - and typically, there are some proven (?) methods of managing large accounts - waterfall and other such methods. However, What we do here is to see how to successfully incorporate Agile practices in managing Analytics assignments. From my experience in both managing web-based applications and Analytics assignments, Analytics is not as "malleable" to following a set of SDLC process. I would love to hear thoughts from enthusiasts on what their experiences are in this. What process do they follow? What has proven successful?

Long Long (Xer)Ago...

We are in the field of customer value maximization. I am not sure if this was by design or by chance, but we seem to be rather well-placed as a hybrid model between an ad agency (though we dont deal with print or TV media), an Analytics consulting company, A technology player and a Process enhancer - We cater to most of our clients needs across these 4 pillars of our company. In short, any marketing initiatives that can be measured is where our capabilities are. We either architect such programs (online brand strategy, usability, building customer segments, CRM roadmaps and so on) or take up currently running programs and help customers derive greater value. While one team is busy designing that awesome creative (incidentally, one of our creatives got short-listed for awards last year), another team is busy help a customer increase sales through his portal, while yet another team is working on building strategic segments of customers.

sample this: http://demo.xerago.com/eps/Portfolio/portfolio/index.htm

Boy, do we live in interesting times!

Web + Campaigns + EBM = $$

I was browsing through Amazon.com the other day, browsing through its various titles, not looking for any one book in particular (though I have been wanting to read Daughters of Arabia ever since I read Desert Royal)

After some 5-10 minutes of general browsing, I switched over to other websites and abandoned the Amazon session. This set me thinking - what if my web Analytics and my Campaign management solution in the back-end exchanged information with each other on my activities on the website?

An intelligent integration of Event-Based Management solution, Campaign management solution and Web Analytics solution can do wonders if done well - I could have got an email from Amazon with a discount on Daughters of Arabia - now, that could have converted me from a casual browser to an actual customer! Ah! Nothing succeeds like success!

Do think about this while I try to check with my local library for a copy!

Value for money!

As we know, most of todays businesses has been hit by the recession (with the exception of pharma, Food industry and a few others). Given the state of affairs, is there even a need to invest on something as (abstract?) as Analytics? The point is yes - it makes more sense to invest on Analytics today rather than in any other time in the past or future. This is because our marketing spends have tightened, we do not want to over-contact customers and end up hurting them, we do not want to contact those customers that would never react to our offers, we do not want to send out fancy, expensive brochures to all and sundry.

The old adage "A penny saved is a penny earned" holds good today more than ever before and Analytics helps us do just that - we can build recession-based models that tell you what kind of books are to be displayed on your shelves to what food can you serve on your airline.

Careful investment on Analytics can prove beneficial to your business (and no matter which business you drive, there is always room for Analytics) and can help take informed decisions on which customers to contact, what to offer them and improve your share-of-wallet overall.

Event-Based Marketing...

What is Event-Based Marketing and how does this differ from traditional Marketing? The write-up below tries to describe EBM and why it can be successful.

Traditional Marketing uses (typically) predictive models that tells us which are the customers that are likely to respond given the offers we have on hand. The predictive model scores the customers (usually, on a sample data set and this is applied on the universe) for further marketing. The flip side (to the traditional approach) is that its still a bet, albeit, a calculated one and it does work well, but many times it tends to go wrong as well. Cut.

Enter EBM, the (old) new kid on the block. EBM is not "likely" marketing or about marketing the offer you have on hand (whether the customer is interested or not), EBM is about watching the customer closely (tracking her transactions) and set up sensible workflows that capture certain events (which are of marketing significance to you) and immediately (well, almost) reach out to her with an offer. The chances of the customer reacting to your offer is higher since the offer is made because the customer has made a certain transaction - no statistical probability here, no predictive models here, no neural networks here. Pureplay transaction-based marketing.

The good thing about EBM is it works! We recently helped a customer design/run some complex transaction workflows and our customers have been pleased with the results they have seen. It works!

If you would like some help with designing Event-Based Marketing programs, do not hesistate to email.

Happy Marketing!