Tag Archives: ncdm

NCDM Notes: Driving customer egagement in the first 90 days

Presenter: Dell’s Global CRM Manager

Dell looked at the actions within the attention -> interest -> new customer funnel to establish triggers.  Initially, Dell was looking at only the conversion step, and felt they needed to also consider other events in the funnel, attention, interest, and post conversion.  If they can keep a person’s interest up, they don’t need to start with this person back at the top of the funnel at attention.

First they build an integrated datamart, with behavioral data, segmentation, contact frequency, sales & support data, purchase history, demographics, preferences, etc.  They are now integrating site traffic data and linking that to offline data.  Sales and support data (mostly unstructured data?) has been the latest to integrate.

Their 90 day program included increasing engagement and experience and increase awareness of other products.  They increase revenue/margin in first 90 days.

The strategy was multi-channel.  It took these 90 day customers out of scheduled marketing efforts.  They provided both marketing messages with educational content.

They generated a 10-40% increase in response and revenue while reducing costs.

So here’s the strategy:

  1. Welcome kit (day 5-7) – mailed (not emailed) and personalized around their system.  Includes a pass-along “refer a friend” coupon.
  2. “Did you forget” trigger (day 14) provided discounts for items that make sense for this customer.
  3. Ratings and reviews (day 21-28)
  4. Solution based triggers (day 29-72) which provided personalized content addressing specific issue of interest to the customer.  They pitched category focused content and items, for example, home network security, home theater placement and products, etc.

Critical to this effort was the creation of a centralized data mart; templates for dynamic content enabling efficient dynamic printing; establishing relationships with global partners.

Execution can be a major stumbling block, given the complexity.  Failure can not be apparent and hard to pinpoint.

Shutterfly Business Solutions:

  • They have the technology to product mass volume but individualized content.
  • They’ve made huge investments in variable content printing.

When working with Dell, they knew they needed to utilize behavioral data and react quickly.  The felt they needed to get materials in the hands of the consumer within 10 days of a purchase, so Shutterfly is focused on getting out materials as quickly as possible.

Shutterfly doesn’t look at dynamic content from the perspective of templates filled with dynamic content.  To them it is all dynamic.  With the Dell effort, they limited themselves to three form factor.

Complicated decision trees are used to build content.

NCDM Notes: Pragmatic Analytics

Presented by Portrait and AAA South

Overwhelming amounts of data combined with rapid growth (especially in unstructured data) makes our job difficult: difficult to gather isight and difficult to implement chages once we have insight.  We also need to make sure we are solving the right problem, and not waisting time on results which are not actionable.

Analysts need to minimize time required to understand and prepare the data: tasks that are necessary before any analytics or modeling can be started. 

There are a couple of solutions:

First, create analytical tables that summarize and aggregate data that is now manually pre-processed.  AAA builds a big table that aggregates data at the household level.  It icludes activity summaries (could be transactional and promotional), prefereces, and  demographics.  For AAA this includes about 150 colums, with some colums updated weekly and some monthly.

AAA uses SAS to create this analytical table, but Portrait could do this too.  AAA is moving towards having Portrait create the analytical table.  This table is kept on its own server that the analytics department controls, unlike the primary DB surver (DB2), which MIS controls.

This more structred data table allowed AAA to help managers do some analytics theselves.

Second, use tools that ease the analytical and modeling process.  AAA uses Portrait software.  The output AAA uses is a matrix of various segment intersections, which simplifies the understand of how each segment will perform; it shows how each element in the matrix will perform relative to the norm.

These two efforts shortened AAA’s modeling process from 5 months to 2 months.

Take-away: Modeling can be easy.  You only need the right structure and tools.

Dinner with Portrait Software

I had the great fortune last night to be invited to dinner by Portrait Software last night.  This is a great group of guys, smart and fun, and if you are considering marketing optimization, analytics, or campaign management software it is worth looking into their suite.

The trip to the Forge, in Miami Beach, was something to remember.  The poor driver had no clue how to get there, so three people immediately whipped out their iPhones to provide directions.  Unfortunately, the iPhone presented to the driver had the wrong address, so we ended up in some industrial area of Miami instead of at a restaurant in Miami beach.  It was a rather hilarious combination of a clueless driver and conflicting GPS advice.

When we got to the restaurant, we were seated around this massive table and in 10 foot tall, 4 foot wide, white leather, wing-back chairs.  These things were so heavy that the staff had to help you push them up to the table.  Once in, you were pretty much stuck.

The chairs also made it a challenge for the great service staff to serve the food and drink.  You”d be talking to somebody and suddenly see a waiter’s arm reach out between two chairs to poor wine, but you’d never see his body.  The contrast between the waiter’s black outfit and the white chairs made this even more comical.

Anyway, the Forge has atmosphere and attitude seeping out of its pores.  I ate too much but drank just the right amount (yeah, right).

Lunch, and one quick note

I had lunch at the “Test & Learn” table.  Although there wasn’t much time for discussion before the presentations, I asked about how people tested offers, given the need to present a public offer on the web.  Only the moderator had an answer beyond coupons and codes and she suggested we try unique URLs, on both side of the test.

NCDM Notes – Target Customers Effectively Through Advanced Analytics

Presented by Intuit & Netezza (a data warehousing company recently purchased by IBM)

The timeliess of integrating data into your marketing process is critical since data ages much more quickly than in the past.

Random note: Intuit is a significant shift from desktop software to SaaS tools, even for their financial products.

Intuit uses the usual variety of channels to acquire customers, but they have found paid ads work best for them.  They also have the usual problems of allocation and balancing spending across channels.  So what data can they use to help with this?  They use DoubleClick to track all converters and non-converters. (Does DoubleClick uses cookies and analytics to attribute customers to marketing efforts. No, it sounds like DC only provides the data, which goes in the Netezza warehouse.)

Omniture provides the data once a person is on their site.  Intuit uses Omniture only to provide data, which it passes to their data warehouse; they don’t use any of the Omniture analytical tools.  Omniture also reads the DC cookie so it can make the link between internal and external data.

With all the data showing what channels (aka media: online ads, organic search, ppc, affiliates, emails) a person was exposed to, they can then do the analysis to allocate the order. 

Intuit also looked at “channel interference” to see if one channel detriments another.  They saw only only a 3% occurrence of an affiliate click happening before a PPC click.  There was more of an overlap in the other direction (ppc click then affiliate) but it was still not at a concerning level.

The key point is that with prospect/exposure level data, they can do the analysis to see what channels or media overlaps.  (This is pretty exciting stuff!)

Behavior of customers is fairly consistent.  If they come in via a paid ad, they are likely to come back in via a paid ad.  If they come in via a specific PPC term, it is likely they’ll come back via that same term.  This suggests there is little cannibalization in the online channels.  (But what about canabilization between online and offline media.)

Renewal efforts at Intuit tanked for a short while when they showed returning customers only high-end products.  When they re-introduced the complete product line, including the free version, response bumped back up.

Since most of the discussion was regarding online marketing and attribution, I asked about their offline efforts and how they dealt with this cross-channel attribution.  The use customer unique vanity URLs (www.intuit.com/victornuovo, for example), which then gives them the data they need to align online and offline data.

NCMD Notes – Text and Sentiment Analytics: Trasforming Call Center, Social, and Survey Data into Customer Intelligence.

I was having a good conversation with the people as Experion, so I came in late to this presentation.

Take data: structured data & unstructured data from social and other external sources; use core functionality to transform that data to some meaningful format; then provide this transformed data to management and analysts.

The transformation of external and internal text type data to meaningful data is very difficult.  You always need to refine the algorithms.  Algorithms need to look at specific keywords that have a certain proximity.

How listening operations fit into DM cycle

Use input from listening services to trigger specific marketing events.  One example, if a person books a room, and mentions that they like a specific Mexican restaurant, then send them a coupon to that restaurant.  (Not sure this makes sense; aren’t they likely to go to this restaurant anyway?  Why give them a coupon?  Better example would be if they mention another, outside Mexican restaurant, then send them a coupon for your Mexican restaurant.)

Make sure you scan and capture internal text data, including data from your call center.  Again you’d need algorithms to transform the text to actionable data.

For example, looking for words “bedroom” and “smell” found in some text string could be a negative indicator. They found that it was important to set expectations before arrival, and then meet those expectations during the stay.

Don’t forget to capture chat text data, and link that to the customer.  Also capture incoming email text.

Monitoring allows you to rectify problems you might not otherwise be aware of.  You need to specifically look at spikes. It is helpful to segment people into promoters and detractors.  If there is a common criticism in both groups, then there is something you need to address. 

Very intersting presentation.  I think it is worth looking into the services Clarbridge provides.

NCMD Notes – Insider Panel: Evolving from DB Marketing to Customer Intelligence

Just a few quick notes, again anything in a blue font are my comments, not those of the speaker…

The key thing when deciding what data to track is to determine what is actionable; there is too much data, especially from the web, so you need to stay focused. You also need to consider the data architecture, so the data is functional.

Chico’s told its analysts that they should spend 20% each week just exploring, pursuing ideas they want to explore.  This is tough, however, because it means turning aside other requests.  (I’d like to second this idea, but it certainly has implications and staffing and systems need to be in place to allow it.)

Analysts need to be able to put analysis into business speak.

Having robust basic level dashboard is critical.  Dashboard reporting can eliminate a lot of ad-hoc requests.  (Not in my experience.  Answering one question, always triggers multiple additional questions.  This means that dashboards need to be dynamc, so executives can drill down.)

There was some discussion regarding the skill sets colleges are providing to students.  The point was made that it is less important to have specific data or statistical skills than it is to have someone who is curious and willing to explore.  Diversity amoung analysts is good.  (This was mentioned above, but I’d add that the analyst needs to have the skills and willingness to share their findings, both in text and verbally.)

“Push marketing is dead in this world”, specifically talking of insurance industry.  Instead, they need to focus on reactive marketing.  “Were pretty good with re-active marketing, putting the customer in the center and providing an esprience on PowerPoint, but not in practice.”

Most of these customer intelligence iniatives are driven from the top, from the CEO level.

Coming trends:

  • Watch out for broadcast 1to1 marketing provided by cable.  This will provide the ability to serve broadcast ads to individuals, based on their unique demographics or other data profile.
  • Social Media
  • Mobile
  • Making data consumers more self-sufficient.

NCDM Presentation: Unlocking Customer Value with Cutting Edge Customer Lifecycle Marketing Approach

I came in late to this presentation, due to a flight cancellation yesterday.  I’m picking up my notes where I jumped in.  The notes may seem a little fragmented, but that’s because I’m only noting down the key points, and I’m trying to do it quickly, so grammar may suffer.  If I add my own comments, I’ll make them in a blue font, so you can distinguish between what the speakers say and my own comments.  I’m not completely sure of the context behind the first couple notes…


Make registration easy.  Don’t load prospects up on survey data, but instead implement a light registration.  Capture limited info early, but be persistent.  More customers that we load into the front of the funnel, the more that comes out the backend.  (I could add some humor that my teenager would appreciate, but I’ll restrain myself.)

Remember, that behavioral data is more valuable than survey data which is more valuable than appended data.  Behavioral data is also live data, you don’t have to wait for it, buy it, or append it.  Appended data can take 30+ days to get into your database.  (This isn’t really true anymore.  There are plenty of vendors that provide realtime customer level data in realtime, using APIs.)

Problems with email: not only that people change email address, but that they have multiple addresses, each with a multiple purpose.  Retailer’s frequently are given a “junk” email account – one that we know will collect lots of spam.  If a customer changes their email address, is it a better address?  (I wonder if people still do change their core email address as frequently as in the past?  I have one or two that are used only for family, friends, and other critical communications, but I have plenty that I use for other, public purposes.  I will replace a public email address if it gets to much spam, but spam filters are pretty good these days, so that hasn’t been an issue in a while.)

Measure open/click rate by age of email address.  It is very likely that engagement drops significantly after 5 or so email.  One goal is to postpone and diminish this fatigue rate.  Make email personal and engaging.


The first 90 days after a 1st purchase are critical.  It can be a good idea to deliver a welcome package via email, or online; it never pays to mail it.  Mailing it can “destroy value”, mostly because it looks like a “mass piece”; they never pay for themselves, even over a customer’s lifetime.  Recognizing the customer with a canned “thank you” can actually hurt your relationship.

There was lots of discussion about making sure any loyalty program is not perceived as marketed to the masses.  Find out what categories they purchased in or viewed and send them offer in that category.

Have planned and tested loyalty program.   Don’t implement it within 30 days; wait till they’re more comfortable with their purchase, at approximately 45 days in their experience.  Note that the 45 days might not be the sweet spot for every company.  You need to experiment and measure loyality program signup rates.

Cementing can be accomplished by quickly getting second purchase.  “Use it or loose it” offers can be “huge”.

Reinforce purchase decision by reinforcing whatever emotion triggered purchase in emails and mailings.  (I think this is key.  If a customer just made a big purchase, make them feel good about it.  There was probably some emotion behind the decision, so leverage that emotion.) He presented the example of BMW.

Customize experience online.  This can provide 60-70% lift.  (Really!?) This can be as simple as the segmentation of products based on purchase or visit history.  It could also be personal area of website with purchase history, product information, ways to use your product, etc.)

One way to personalized online content is with the use of ad serving software, which is (supposedly) pretty easy to plug into most websites.

Make order process as simple as possible.  The goal is to make it simple for a person to re-order.  For example, pre-populate any fields that can be pre-populated.

They provided a grocer example: most grocers (even the big national chains) do an awful job of leveraging the massive amount of data they have.  Tesco (UK) is the exception.  Their offers lead with things they know you buy from them, but also include products you don’t buy but which are good cross-sell opportunities.  This way they try to shift some of your wallet share to them.

Try merging primary consumer research data with behavioral data.  Customize offer based on “potential” value instead of “historical” value.  Quote: it is very hard to determine a customer’s potential value.  You need to know wallet share.

Watch out for silent attrition.  Listen to triggers, changes in usage patterns.  He provided an example of a bank.  Instead of watching for closing of accounts (when it is frequently too late), watch account balance, number of transactions, etc.  Declines will help determine who is going close account.

Some products are different.  With insurace, the last thing you want to do is remind them they are a client.  It tends to remind them to shop around for insurance.

Think about what other kinds of rewards can be presented. (He mentioned foursquare, as an example.)  Is there some form  of recognition that would reward the customer.

Loyalty programs – things to track:

Watch out for loyalty programs destroying value.  It may just give better deals to people who would buy anyway.  Loyalty programs rarely provide direct benefit for this very reason.  Value is in capturing data that helps you market, especially if that data is sold.

It is very tricky to measure incremental value.  (No kidding!)

There was some discussion on “share of wallet”: only way to get at this is metric is through a survey.  Need to ask them where else they shop, and how much.  (I think even the survey approach is limited in its value.  How many people are going to tell one vendor how much they spend with other vendors?  How reliable is that data when provided?)

Customer Level Data/Analysis

It is important to measure EBITDA by customer segment: new, spent less, spent more, lapsed.  If you know where your EBITDA is coming from, then you determine how to allocate your marketing budget.  (There were some great charts and graphs.  I’ll have to see if I can find them online.  If so, I’ll post.)

Also measure $GM by various RFM segments or other metrics (# channels used by customer, for example).

Look at various marketing drivers (cross/up sell efforts, frontline sales, loyalty, marketing, etc.) and the value they provide.  (But that raises the issue of order allocation – crediting)!  Best way to get at this is by survey data: by asking what triggered purchase.

Always “test & learn”.  Test everything you do, otherwise you don’t know.  That’s the one thing to takeaway, if we take away nothing else.  This is the biggest lever you can pull.  Learning is the key element.  Need to be able to quickly implement results of test.

This usually involves tracking results down to the customer level.

Always include “hold-out” samples that helps you evaluate mailing efforts.  (This might be difficult for vendors with high ticket/low conversion products.)

Cutting edge approaches to maximizing customer value take years to implement.  For that reason, long-term executive commitment is critical.

Final comments…this presentation was crowded, or maybe the room was just too small.  I got a seat because I brought one in from the lobby and placed it in the back.  One person, who came in later, sat on the floor.

Overall, a good presentation.  Takeaways for me:

  1. Test & learn, which I already knew.
  2. Make registration process easy, and incrementally capture data.
  3. Really think about the actions within 90 days of a first order.  This is a high value group to target.
  4. Need to focus on making emails valuable to the readers, so they won’t fatigue and so the prospect will give us their most read email address.

At the NCDM show in Miami

I finally made it into Miami this morning for the NCDM show.  I was supposed to fly in last night, but high winds kept my plane on the ground.  I ended up taking a 7:15 plane out this morning, got in at around 1PM, and got to the show by 2PM, where I walked into the middle of a presentation.

The NCDM show is a lot different than when I last attended, back in 2005 or 2006, I think.  The big difference?  It was a lot bigger back then.  They pretty much took over the Gaylord Palms Convenction center in Orlando back then.  Now, I would guess that the floor space is a quarter of what it used to be. 

Still, the important part is the ideas it generates.  I’ll post occasional updates, especially of presentations.