I.R. Gilyeat & Company - Our Latest Thinking
http://blog.irgilyeat.com
I.R. Gilyeat & Company - Our Latest Thinking

Marketing automation - and an over-dependence on email address

Pardot has an over-dependence on the email address...and I suspect other marketing automation vendors do as well.

Following me along in this very common business scenario:

Company "A" is uses Salesforce.com and they have a few thousand prospect accounts and associated contacts in their CRM system.  They have a pretty complete Contact record with the exception of email address.  However, their inside sales team is disciplined and they capture an email address each time they call and talk to one of these prospects.

Guess what?  Pardot cannot automatically download this Contact into their marketing automation platform.  Really?  Yep.  Confirmed this with their tech support team....

The basic fact is, when a Contact or Lead record is created in Salesforce, it MUST have an email address or Pardot make a match.  In fact, if the records was created first, then the email was added and Pardot makes a match, that record will have been flagged by Pardot as "Do Not Sync."  Which means, if you want to trigger a message to a prospect or drop that prospect into a Lead Nurturing stream (the bread and butter of marketing automation) you cannot do this with an automated solution in Pardot - IF the email address was added to the record AFTER it was created.

Here are the two solutions that Pardot offers:

1)  Manually do an export of records from Salesforce and manually do an import into Pardot.  This will setup the sync so that Pardot can notify your CRM it is ready to download the new information - based on a match to the email.

2)  Manually press a "Send to Pardot" with the CRM integration tool set - which means your company is entirely dependent on the inside sales team to set that marker and drop the person into a nurturing stream.  This isn't entirely bad, since it gives control to the sales team, but you definitely have a gap in the "marketing automation" equation. 

As you can see from the above example, there are still gaps in aligning the worlds of sales and marketing.  Progress is being made, but keep your eyes wide open and keep your expectations in the realm of reality.  Marketing automation is another useful tool - but it's no silver bullet.



A new notion on messaging security...

Generally speaking when we think of security today, each person, like you and me, sets up a password, creates a series of "challenge questions" and maybe even identifies whether the device being used is "public" or "private".  In essence, the security information is given to the system that you want to log into.

Perhaps we have it all wrong...

In the future, with the emerging abilities of apps, big data, cognitive computing and mobile devices, the device itself will be the "giver" of the password.

What does messaging security look like if "no one knows the password except the person to whom it is given?" - with emphasis on "given".  Perhaps it is the message or the system that must become intelligent enough to "give" a password or access to each of us - in the moment in which we need it.

A few years back, I worked for a company that used key fobs to randomly generate a password every time I wanted to log into the system from a remote location.  It was a bit cumbersome and I know the technology is still in use, but perhaps our litany of mobile devices and access points need to incorporate this same basic notion:  the password is known only the person to whom it is "given."

Dear marketers and creators of digital messaging platforms: ExactTarget, Responsys, Pardot, Unica, ConstantContact, Eloqua, Marketo, Act-On... and so on down the list... as messaging systems become extremely personal, our messaging systems must keep pace and security will become a neccessity - you know "table stakes" for entry into the game - and not just for access to the "system" but for access to each and every message sent and "received."

Lead scoring limitations...

One of the operational aspects you want to understand when choosing a marketing automation platform for lead scoring is to know the limitations of the environment.

As an example, will the lead scoring engine allow you to build multiple scores for different segments of your customer base?  Or is the scoring engine one dimensional as in all scores are applied equally to all records?

Let's say I am marketing to the healthcare industry.  I want to create a lead score for the executives, a different score for the CFO or P&L owner and a third score for medical personnel such as nurses, anesthesiologists, surgeons and so on.

In our experience with Pardot, it is a one dimensional platform - all behavioral scores are applied equally to all Prospect records.  True there are "Grades" that can be applied to specified attributes like Title or Number of Employees, but the ability to create multiple scores for differing segments of the file is non-existent.  When we asked this question to Pardot, they agreed with our assessment.

Marketo on the other hand allows you to create multiple scores:  Composite score, product score, demographic score, etc.  This is a much richer environment but it can also be much more expensive.

Bottom line, know what you need and what the limitations are before you buy.

Too much innovation in marketing automation?

Is there too much innovation happening in marketing automation?

The question above shows up for me because I just finished talking with an equity analyst and his boss the VP of equities - this was in a multi-billion dollar (is there any other kind...lol) firm that is deeply involved in the technology sector...

The thing that strikes me about this conversation and several others I've had like it recently, is their approach...they are looking at these companies like they are software companies.

In my view, this misses the mark...

The marketing automation platforms of today (Marketo, Pardot, Eloqua, Unica, etc.) are already robust enough to do everything that 80% of marketers want to do.  The marketers are not lacking in technology.  There is so much technology already available, that it is overwhelming to many marketing teams.  Yes, the 20% of the market are crying for more sophistication and more capability... and they're driving the innovation market.. but from an investment standpoint, the place where you get scale...I would argue the market has sufficient functionality to satisfy most marketers and that its mostly about how these companies market themselves and has little to do with functional comparisons between the vendors offerings.

Simple, simple, simple is what the market needs.  Simple answers. Simple solutions.  Simple implementations.  Simple execution...  

www.irgilyeat.com

Perfect Recall – There’s an App for that…

Customer life cycle marketing reaches a new level of “customer intimacy” when the Smartphone uses cognitive computing to enable perfect recall.

Consider the marketing requirements when the Smartphone uses DNA sampling and near-field wireless communications to scan, analyze and enable perfect recall from the human brain…

…swipe your finger across the screen…and the “tactile sampler” uses your DNA to “unlock” the records in your mind.  This includes everything you have ever seen, heard, tasted, touched, felt or imagined.  The record has everything you have ever said, thought or experienced.  
The DNA sample is the golden key that identifies perfectly that it is in fact “you” and opens the app to scan and retrieve everything stored in your mind…

The perfect recall app is a service of “IBM” and is part of their “Watson” cognitive computing service…

The interface is your voice…once the app is open you speak what it is you are searching for and the cognitive solution retrieves the past memory from your mind and displays it in full color, rich media on the screen of your smart phone… everything is there… the perfect recall function of your mind has been opened through the DNA sample and the Perfect Recall app that you downloaded to your Smartphone device is doing its job…

…You can even choose to “broadcast” the experience and share it with your friends and family that are with you…not that they’re necessarily physically with you…but using telepresence capabilities and free-form optical network display technology it is played out in “living color” right before your eyes and their eyes.

 The rudimentary technology of Google Glass was useful in its time but clearly unable to match the capabilities of the two eyes already implanted in your head and integrated with the perfect storage capabilities of your brain… however, as a “playback” device it has proven to be useful as it provides the “device platform” for “customer life cycle experiences”.

Remember what I said in my earlier posts…

There are no secrets on the web… and everything is on the web…light is the law which governs all space…and there is no space in which there is no law.

So what does a marketer have to do in this future realm of customer life cycle marketing?  Remember, agency is always present, or in the vernacular of today’s marketer, the customer has “opted-in”… and technology is simply an enabler of choice.  

When the marketing department includes a genetic scientist and the consumer is using a DNA sampler on his/her Smartphone; when this technology is enabled with cognitive computing services, relationships of trust are still required.  Whether those relationships are person-to-person or between a person and a company; the relationship still must be built on principles of mutual respect and a desire by both parties to stay in the relationship.

If my Smartphone uses a DNA sampler and I trust my optometrist – which I do, with certain parts of my physical experience – then if the DNA sample taken today through my Smartphone indicates a developing problem with my eyes, an “automated alert” message can be delivered to me in real-time asking if I want to schedule an appointment with my eye doctor tomorrow morning at 10:30am – the doctor happens to be available due to a cancellation today…

In addition, if “Allergen” is a trusted partner of mine, I want to retrieve an offer from them for their newest contact lens that was built using cognitive models; their new product “reprograms” the neurons and chemical substrates in my eyes to prevent the eye problem that was diagnosed through the DNA sampler…

Sounds farfetched?  Not really…the technology is coming and most of the rudimentary capabilities are already in market.  Marketers really do need to become experts in customer life cycle marketing and messaging platforms need to continue to scale up... and of course companies need to build relationships of trust.


Marketing to the "segment of one"

"The size of customer segments is shrinking down to one... " so says Oracle in their report released earlier this year:  “Global Insights on Succeeding in the Customer Experience Era".

I suppose this should surprise us?  No, not really.  Peppers & Rogers founded a company on the whole idea of 1to1 marketing.  "Personalization" in marketing has been sold as the golden key for a long time: remember the idea that changing the letter from "Dear Friend" to "Dear John Doe" would give you greater lift in your direct marketing; and of course we needed to personalize the outside envelope from "Resident or Occupant" to our first name and last name.  American Baby, some 20 years ago was creating personalized magazines down to a "unit of one" with full variations in advertising, editorial content and insertion of the new mother's name inside the copy of the magazine article; personalized books abound from Flickr and others.

No, personalization and marketing to a segment of one is not a new idea.  The technology may have migrated from the mailbox to the smart phone and the time frame of message delivery has compressed from the printing press to the web storefront or from the "inbox" to the mobile device... yes, we are compressing the time frame of delivery...but the fundamental concepts of personalization haven't changed all that much...

We still want to be sold to and buy things as individuals...

So what's a marketing pro to do?  Learn from history and apply past customer experiences to the present technology.  Example:  if your marketing has "lists" in it and you're building, managing and using "lists" in your marketing automation processes, get a copy of the following and digest it and implement it for the next few years:  "The Completed Direct Mail List Handbook" by Ed Burnett.  It contains "Everything You Need to Know About Lists and How to Use Them for Greater Profit."  This is a 762 page tome with detailed operational know how on how to squeeze profits from "lists".

...but doesn't this conflict with the whole concept of personalization and marketing to a segment of "one"?  Well not really... 

I'm a human and so are you.  We all fit into that bucket or segment.  I'm a man and so are some of you... well, now we have two buckets.  My DNA strand is different than yours... yes, well, I have a mother and father and so do you.  Another bucket or segment. (Ancestry.com would be proud.lol ).  I buy clothes and so do you.  I drive a car and so do many of you.  I eat food and so do you... my point is, the segment of one is real, but don't lose sight of the fact that everyone belongs to some kind of group or dozens of groups / buckets/ clusters / etc. that can be useful to managing your marketing operations.

Marketing to a segment of one is here to stay, but then again it always has been... 

"Follow the sun" messaging solutions...

A few years back I worked in an outsourcing firm where we developed and provided "follow the sun" logic for fulfillment of ecommerce orders.  The simple realty was it was cheaper to ship a package from a warehouse in Orlando to a customer in Miami than it was to ship a product from Los Angeles to a customer in Miami.  However, it was equally as important to set the right expectation with the customer that his product would "ship today" and that answer might be different if the Orlando warehouse was closed and the product wasn't available from the Los Angeles warehouse.  Basically, warehouses would be turned off and product would become unavailable to ship that same day as the sun moved west and warehouses turned off for the night. As you can imagine the location of the customer, location of the product and the order time frame all had to be coordinated in order to set the right expectation with the customer.

Amazon is basically following this same model with their recent partnership announcement with the U.S. Postal Service and their latest developments in the use of drone technology - yet to be brought to market - but being able to set the right expectation and deliver "same day" instead of "next day" is the frontier they are chasing with drone deliveries.

Now let's consider this concept of "follow the sun" logic in messaging systems and let's apply it to "follow that storm" instead.

Imagine with me if you will our ability to deploy a drone that is programmed to send an alert in real-time based on the location of a storm.  First, off it will need to be very durable - but okay, we can solve that.  Second, it will need to contain "an embedded device" that is plugged into a wireless messaging platform.  Third, the algorithms will need to know when to trigger the message and will need to know who to send the message to.  Those aren't necessarily difficult challenges to solve. Today we have robust messaging platforms that can send millions of messages per hour to individual recipients.  However, what we need here is the ability to send millions of messages per second and the message needs to be triggered by algorithms and logic only to those people that have "opted-in" to receive the alert....

Now consider triangulating the following elements:  location of the storm, destination of the package(e.g. where is this storm going to be delivered) and speed of delivery (e.g. overnight, same day, in the next few minutes, etc.).

All of these elements can be pulled together into a wonderful messaging solution that delivers alerts to residents within the storm track based on opt-in decisions by those that live in storm prone areas.  Yes, it will require some smart logic, data tracking, hefty computing power and an industrial strength messaging system to send a million messages a second - but this is where we're going and it's not too far off.

Now how many other uses can you think of where "follow the sun" or "follow that storm" logic can be applied?  Here's a few to consider...

Hurricanes, tornadoes and tsunamis.  What about forest fires?  Or freeway accidents that cause massive traffic jams or pileups?  Snowstorms for skiers that want the freshest powder and surfers that want the best waves.  Delivery of produce or movement of herds for the outdoorsmen; movement of game fish in the ocean and migrations of herds on the savanna.  Or maybe just maybe, applying nano-technology we can track and avoid the movement of viruses and pestilence across the planet.

As I've said before in my posts, there are no secrets on the web...and everything is on the web.

Frankly, the possibilities for messaging systems are enormous and we're just getting started...


Off-grid - outside of the pipe - free space optical networks

"Big data" is an idea that fires the imagination.  Combine this with the ever present "smart phone" and "free space optical networks" and perhaps we land at something I'll call "persistent communications" or maybe more appropriately "big ears" or "big eyes" or "big brother." However, big brother conveys the intrusiveness of someone always watching and what I'm getting at s the ability to "always listen" or to "always communicate."  One company I'm familiar with - Avnet - refers to it as "constant connectivity."

Next March in Berkeley, California, The Economist is holding an innovation forum.  The theme for presenters is something along the lines of, "Are we asking big enough questions?" and "How do we get there?"  In response to these two questions, I'd like to suggest the following as a framework:
  • There is no space in which there is no law.
  • Light is the law which governs all things.
  • There are no secrets on the web.
  • All things are on the web...
  • Mass x velocity = momentum
When I think of constant connectivity, as Avnet refers to it, I think of light and how soon will we advance technology to the point where we are in effect "off the grid" or "outside of the pipe."  And what I mean by this is how soon will we move beyond the limits of fiber optic cable?  If we intend to have massive amounts of "big data", with constant connectivity, then the limits of fiber and the short-comings of current wireless networks will have to be overcome.  However, once we figure that out and we begin to work in "free space optical networks" then the physical limitations currently imposed by our ignorance will make constant connectivity pretty exciting.  

I am suggesting that light is the law that governs all things and there is no space in which there is no law.  Assuming these two points to be correct then there are literally no secrets on the web - all things are on the web.  It's only a matter of how that web of constant connectivity is governed and used.

So do do we get from here to there?  Well, Avnet and other companies like them - meaning those that have global scale - are in a grand position to help.  Device level, embedded components with adaptive & predictive messaging systems (think IBM Watson scaled up with ExactTarget or other digital messaging platforms) and I think we're on the cusp of some great inventions - and massively persistent communications - even as persistent as sunlight itself - and all things (devices or things that respond to sunlight) that respond to sunlight.

Now go back to my recent posts on self-directed segmentation models and I propose we are a lot closer to fulfilling Google's corporation mission than we realize.  There mission is to organize the world's information and make it universally accessible and useful.  Now I don't think this will happen tomorrow and this doesn't mean all things will be visible and accessible to all people... 

...however... and this is where my last point comes in...simple physics teaches us that mass x velocity = momentum.  Momentum implies power and impact.  The larger the physical body and the faster it moves, the greater its impact.  Apply this concept to billions of people with increasing levels of education and wide scale access to massively organized "big data" and we will see innovations accelerated in ways that will be monumental - in ways that we can scarcely imagine.

Self-directed segmentation and the divisions of humanity...

Earlier this month Rob High, CTO of IBM Watson Solutions, posted on Forbes the following: "Cognitive machines will democratize expertise."

This is a broad statement, that could be applied to all of humanity so let's consider it in the context of self-directed segmentation models...

Self-directed segmentation models suggest that cognitive computing, similar to capabilities offered by Watson, enable us to build models that self-adjust; models that "ingest data"...like Watson, recognize new attributes and have the ability to apply context to those attributes; add the new attributes to the model and adjust accordingly.

When this concept of self-directed segmentation modeling is applied to all of humanity, the idea requires some kind of massive input source.  Today, that input source could easily be the widely used "smart phone."  Smart phones can be carried by billions of people and as they continue to advance the limits of data capture at some point DNA sampling can occur - in real-time, continuously, and wallah we have massive sources of "big data." Consider also the attraction of "life loggers", those that wear cameras to capture every moment of their life; a billion+ deployed cameras around the globe; trillions of Twitter messages, social posts, etc. it's not to hard to imagine enough big data input sources for cognitive computers like Watson.

Simply said, we are on the cusp of having sufficient data to model the behavior of all humanity, even at the individual level across a lifetime from conception to death - and what I'm suggesting is this modeling will be self-directed by those who willingly provide the data - and in some cases by those who unwillingly participate due to government mandate  (e.g. traffic cameras or health directives, etc.).  In short, there will be no shortage of data.

Back to the statement, "cognitive machines will democratize expertise"... without diving too far into the discussion of democracy and what it suggests about people's behavior, let's consider the following instead as a contrasting point of view:

Democratization of knowledge and expertise is a grand idea that disintegrates under the harsh impact of mankinds' agency.

Agency is the ability of each person to think and act for himself.  It is the ability and initiative of people to grow and expand.  Unfortunately equal responsibilities, equal intelligence and equal effort is not obtained - even when equal opportunities are available.  As a result, many privileges and benefits are lost even though they are readily available to anyone who is willing to work for them and master the laws and rules associated with those privileges.

When we combine the "smart phone" + agency + big data + cognitive computing I believe we end up with self-directed segmentation models; even a grand separation of humanity based on the agency of each person.

Technology is marvelous and grand inventions are happening all around the globe.  However, we must remember, technology is simply an enabler of choice - even in the grand scale of cognitive computing - and especially when we apply it to expertise.
 

Self-directed segmentation models are coming...

Imagine if you will what happens when a smart phone is combined with the capabilities of Watson...remembering that a core operating tenet of Watson is to "ingest data"...what you eventually arrive at are self-directed segmentation models.

Self-directed segmentation modeling is "clustering" at it's finest.  In the old days of database marketing demographics were applied to lists of data in order to enhance them with things like household income, age, gender, education, type of vehicle driven, etc.  It was exciting to build models with hundreds of variables applied to millions of records.  In those days we were thrilled to see lift rates of 50% across the entire file... but let's put that in perspective - this was on a base response rate of say 1.6%.  So moving it to 2.4% was significant. 

Now consider what a self-directed model might accomplish when the model is built on data "ingested" into IBM's Watson by an app that you download onto your smart phone.  In today's world that could be a lifetime of data willingly provided by "life loggers" that wear cameras on the clothing and record every moment of their entire life.  Or it could mean narrower amounts of data gleaned from Facebook across 1.2 billion user accounts.  Or perhaps it is a data set given up willingly by you from your cell phone so your doctor can help you respond to the challenges of diabetes, Alzheimer or any other physical affliction.

So, why are these models self-directed?  It's because a process is defined by which the model ingests data and uses cognitive computing to adjust and evolve over the course of time based on the data that is ingested.  The model is not built based on a pre-defined set of attributes.  The model accepts all data ingested and places it into contextual understanding based on all data that has been ingested to that point in time... and continues evolve over time - in effect the model is self-directed based on the massive amounts of data that is constantly consumed.

The possibilities are remarkable and the benefits can be many.  However, let me restate what I have been telling my children for many years - there are no secrets on the web - and let me add, everything is on the web.

Blog Software
Blog Software