Candidate CRM & Workflow systems for Data Driven Marketing

I thought that I would share some ideas that we have been working on in the areas of Sales/Marketing/CRM… We are calling it Data Driven Marketing, and it is all about making it easier for our customers to use our products better, and to use more of our products.  We are looking around for a platform to base this on, and a CRM with workflow (maybe an add-on), and good external systems integration capability seems to be a starting point.

Some requirements

  • Ideally Free Open Source Software, so that we can quickly deploy it and manage it
  • Available as SAAS (Software as a service) or as something that we can install on our servers
  • A graphic flow designer, so that our sales, marketing and customer teams can plan the experience directly
  • Can “listen” to customer events (e.g. open email, click email link, system reports customer missed a phone call)
  • Can communicate with the customer (e.g. send email, put onto a customer call list)
  • Can talk and listen to our other systems
  • Gives us access to the CRM data within the system
  • Good reporting, so that we know how the flows are going on, maybe with A/B comparison

Some candidate systems for Data Driven Marketing

Hmmmm…Here is a little experiment that I am working on, I thought that I would do the comparison and research in a google docs spreadsheet, and publish the result on the web. It is very much a bit of work in progress, so please forgive the “alpha” nature of this



CRM Application WF Application Cost
basic co-ordinates Functions
process Manager Enterprise 2.0 SierraCRM USD 499 Once BonitaSoft FOSS process Manager Enterprise 2.0 ProcessMaker – SugarCRM Edition FOSS + Ask

How MoNo-MeLi could rule the world

My recent blog posts, “Does LAMP need a successor?” and “Introducing MoNo-MeLi – the heir to LAMP’s thone” generated quite a bit of conversation on LinkedInOff the back of these comments, I’ve managed to develop the MoNo-MeLi plan further.

So, the key quality criteria for using a platform to perform a certain task are:

  1. Ease of getting started
  2. Availability of hosting, tools, etc
  3. Complexity of purchase/acquisition
  4. Quantity of people familiar with technology elements
  5. Production efficiency
  6. Operational efficiency

Criteria 1-4 relate mostly to the platform itself, and here LAMP scores highly:

  • Linux in its purest form is not that easy for most people to get started with. However, if you have Linux as part of a SAAS package, the complexities are masked. People are not using windows, because it is not free, and it can be complicated to purchase
  • Apache is easy to get started with, but there are other good web servers around.
  • for MySQL, lots of people know a bit of SQL, and purchasing is simple (providing Oracle don’t disrupt things too much)
  • PHP¸ while many people might regard it as primitive compared to, say, Ruby on Rails, is well known, and easy to get started.

For criteria 5 and 6, it depends more on the task being undertaken. In my view there are 3 different types of web content:

1. STATIC which can be served by a simple CMS system, or straight HTML

With static, generally people are not using LAMP directly, though it may well be included in their SAAS package (such as WordPress). LAMP scores highly on production and operational efficiency, especially with modern WYSIWYG editors (which may make one believe that there is a real-time element).

2. MODAL/TRANSACTIONAL which will adapt, given circumstances, and allow forms to be filled in, and can be served by PHP, Ruby, etc.

It is in modal content that LAMP starts to show signs of strain.

3. REAL-TIME, such as chat systems, Google Drive, etc, which is harder to serve with these platforms.

This is what got me started looking for LAMP’s successor. Linux is ok, but Apache is far too hungry. MySQL, while good on the transactional and reporting elements, yields to Mongo’s superior real-time skills. And for PHP, currently it’s plain old hard work at the moment to make it real-time.

How could MoNo-MeLi rule the world?

What would it take to get MoNo-MeLi in line to inherit from LAMP?

  1. Get some tools that use this platform into wide-spread use, e.g. CMS, shopping system, social system, simple transactional frameworks.
  2. Make it incredibly simple for people to get started, so they will prefer it even for tasks that would traditionally use LAMP.
  3. Find some compelling real-time elements that can add value to existing ideas.
  4. Make the transition gradual. This way people can adopt MoNo-MeLi easily without the pressure of a sudden move.
  5. And the alpha and omega – apps will make or break MoNo-MeLi!

This thought process was driven by the responses I received to my previous articles, soif you have any more comments for me about this post, I can’t wait to hear them!

MoNo-MeLi as the successor to LAMP: responses

I’m back from a short break away and have been enjoying reading some comments people have made about some recent blog posts of mine. You may have read them, “Does LAMP need a successor?” and “Introducing MoNo-MeLi – the heir to LAMP’s thone”. If not, this blog post will probably make far more sense if you do go read them.

Since writing them, I have had some very interesting feedback, representing a broad spectrum of perspectives. Here are some of the highlights:

Firstly, in the node.JS developers groupAJ ONeal discusses the distinction between the different classes of web content, and postulates that even if MoNo-MeLi can’t take on LAMP itself, it could help out with non-CRUD tasks. Douglas Muth talks about some of the many merits of node.JS


In the Meteor group, a few people made some very relevant comments – Matt Cameron in particular thinks that it’s the apps that will make or break MoNo-MeLi.

Finally, the biggest discussion was In the Open Source LinkedIn group.


Malcolm Simpson believes that since LAMP has such a large community of users, it will surely continue to improve and evolve, rather than being replaced. Mark Kimsal mentions that early user success is vital. Michael Grant points out that PHP is successful due to its ease of use. Laurent Caillette adds some really interesting comments on the various alternatives to PHP. Dave Howe asks some important questions on the migration, and discusses the issues of adopting too early.

I’m very grateful to everyone for their contributions, and check back tomorrow for a blog post building on these points… how MoNo-MeLi could rule the world!

dmSwitchboard12 Version 2, what will the Ghosts make of you?

I recently introduced you to my Ghosts of Product Managers Past, Present and Future. Now here’s a problem for those Ghosts to solve.

switchboard2We have just released dmSwitchboard12 Version 2. This is our hosted switchboard, which is universally agreed to be the best in the world, (from a sample base of myself and some people whose livelihoods are dependent on it).

In September 2012 we launched the first version: a “revolutionary” release, i.e. lots of new features (by my count it included more than 20 world firsts!). Since then various features have come in as patches, and with this version a shed-load of “evolutionary” features (i.e. not world firsts, but building on the successes of the first version to keep our users happy).  It’s in preview, meaning that “early experience” customers can use it, but it is not being offered as a default for new customers, and there is no sunset of v1 for the existing customers.

What did my Ghosts think of this? Remember we have :

  • Bill, Ghost of Product Management Past,
  • SteveGhost of PdM Present, and
  • I (Rob) will play the role of the future Ghost, for argument’s sake, just to get them talking.

Steve: The most important thing is to keep the product looking simple.

ghostsBill: No, no, add loads of features which have been requested by the users!

Rob: Remember, guys, we want to lay foundations for future releases (particularly “Unified Business” integration), with some taster features.

Steve: Why not just surprise them with something really cool?

Bill: That works for me!

So, having brought my Ghosts to compromise, we have:

  1. introduced layering of features
    • only essential and simple features are in the immediate layer (Steve’s plan)
    • the rest go under “advanced options” (Bill’s preferences included, so no one at Microsoft throws any chairs)
  2. given people a clear path in and out of the new version. It’s important that any upgrade is not one-way (we don’t want a situation reminiscent of William Hague’s “the euro is a burning building with no exits”), so there should be an upgrade AND downgrade option. It’s one thing if your new version of Android does not support “Zombie Super Pigs”, but switchboards are part of company’s business process.
    • Steve: Make the product so perfect that you don’t need to give them an out!

And before we go to the live release, we will:

  1. do some gardening on the glossary tree: terms are really important (Steve approves)
  2. create some videos and planned journeys based on user types and typical cases (Steve wants us to keep them simple)
  3. not try to make it as minimalistic as Apple does! One needs to spend millions of pounds to get products of this complexity looking swish (a lot of companies try this and get it very wrong!)
    • Steve: At least you’re admitting that you’re not as good as me, Rob!

For all the product managers and would-be products managers out there, why not borrow my Ghosts and give them a grilling the next time you come up against a particularly tough PdM problem.

And now a word from our sponsors…

I, Rob Darwin, hope very much that you have enjoyed this blog article (and I’m sure the Ghosts would concur)!

If you did, why not chuck out your old switchboard, or PABX if you have one, then rush down and buy one?

…in fact they’re such fun, why not buy two?

The Ghosts of Product Managers: Past, Present and Future

I’m a product manager, amongst other sins, for a telco on London’s “Silicon Roundabout”, and sometimes when one is wrestling a product out the door (you can see the finger-nail marks around the door-frame, as it resists your attempts to release it), one wonders about product management.

Product Management (PdM) is about choosing where to compromise and where not to.  Steve Job’s biography by Isaacson never seemed to get quite to the heart of it (from my POV), and maybe no one ever will.  But Steve is a useful ghost to have around.

Bill Gates shall be my Ghost of Product Managers Past: the techie driven federation of features all jumbled in together.

Steve Jobs, then, is the Ghost of Product Managers Present, the highly curated Appie approach.  Here the user story is king, and everything looks very minimalistic. The problem is that you can spend hours looking for something important, as clear signalling is not as important as a “cool” appearance.

And who is the Ghost of Product Managers to come? Bill is retired and spending his billions on good causes, Steve is dead… so who comes next? Both Microsoft and Apple are ever evolving, but Google could be a strong contender with its Android OS. Who’s your Ghost of Product Managers Future?

Later this week, I’ll reveal how the Ghosts helped me to design one of the most technologically advanced pieces (in my humble and not at all biased opinion) of telecoms equipment on the market.

Introducing MoNo-MeLi – the heir to LAMP’s throne

Last blog post I posed the question, does LAMP need a successor? LAMP, a software bundle comprised of Linux, Apache, MySQL and PHP, currently powers the web, but due to changes in the way consumers operate online, it’s swiftly becoming old-fashioned and too slow for the real-time, cloud and app lovers of today’s internet, not to mention expensive to code for, given the today’s current expectations.

Meet MoNo-MeLi.


  • Mongo replaces MySQL as the database of choice – less of a reporting tool, more of a data/document storage system, but with some nice real-time characteristics.
  • Node.js is the new mainstream platform, stunningly fast and good at multitasking.
  • Meteor is a small start-up in San Francisco, not even in version 1, but has real potential to tie everything together because it makes it easy for the web to go real-time. Have a look at some of their demos for a quick example.
  • And Linux is still in the frame, upping its game to stick with the acronym that powers the internet. Node.js is taking an increasing share of its workload, though.

We threw around a few different acronyms here – LMMN, nSLaMM, nJMM – and MoNo-MeLi (moh-noh-mee-lee) won out in the end.

It builds on the successes of LAMP, while offering improvements in a lot of the areas where users are demanding better facilities:

  • It slims down the transactions between web-client and server to a bare minimum.
  • This means applications don’t have to reload the whole page…
  • …and can load in as little as 100 milliseconds.
  • The programs require fewer lines of code, meaning shorter development times…
  • …and compared to some of the other up-and-comers, there is a lot of online support  for these packages.

You heard it here first. One might say LAMP is on the way out, as it does not offer a High Productivity Real-time web. But MoNo-MeLi is ahead of the pack, eager to seize its throne.

Does LAMP need a successor?

At Digital Mail we are working hard on implementing a new architecture for a product that we are working on. We need seriously real-time software, and lots of simultaneous connections and activities.

As developers know, the web is powered by LAMP, a software bundle comprised of Linux, Apache, MySQL and PHP.

However there are some trends that are putting this dominance under threat:

  • Desktop applications are being replaced by cloud/hosted apps, and it’s hard work to develop this in LAMP. Something that removes a lot of the web-client and server separation coding pain would be ideal.
  • With more interactive “application like” websites, if the whole page has to be reloaded whenever something is changed, that’s a lot of refreshing.
  • People like responsiveness and know that they can get it on certain web sites that are out in the market, so aren’t tolerant of slow apps. One should be aiming at responsiveness of 100-250mSec.
  • There is a shortage of good developers, and the good developers need to be even more productive to support competitive timelines in ambitious roadmaps.

As I hope you can see, we think that LAMP does need a successor. Later this week I will outline what we propose to be the successor (we call it MoNo-MeLi) and why we think it solves all these arising problems, so check back soon.

On seperating Operational, Tactical and Strategic Computing

I was recently involved in a discussion on OpenERP, when I suggested that one might export data from the main ERP system for seperate processing using SQL and Excel for reporting.

This was regarded by some as a criticism of the ERP system’s reporting capability, however I think that this is quite natural for the following reasons:

  • Even if it was necessary for a company to use special SQL and Spreadsheets, I do not think that it would render OpenERP useless
  • There is the classic 3 layer model of IT in companies: Operational, Tactical and Strategic. OpenERP is predominantly an Operational & Tactical system, with a strong mission to provide a reliable support at the Operational layer. Strategic systems tend to be very different, they have to be much more fluid, but they do not need to be operationally reliable.
  • For strategic and high level tactical work, one tends to need a very fluid system, which does not need to run to operational standards. Excel is very good at this.
  • Strategic modeling tends to have very short development cycles of the model, which can be iterated several times per hour
  • Strategic modelling output should only be fed back into the operational model with a great deal of caution, as it tends to be pretty “hacky”

This was why we separated our strategic work from our live operational systems. We take an overnight copy of all our databases and then do certain tactical reports (which take about 4 hours on other machines), and export stuff for strategic work in Excel.


CRM and “the one post-it note at a time” syndrome

Imagine, the phone rings, and as you take the call, a colleague hands you an A3 sheet which is full of useful post-it notes for the call:

who the customer is

their contact details

key details of the last 5 calls

the product that they are thinking of buying

and the 5 key sales tips.

Wow, that’s a good way of starting the sales call!


Meanwhile on plant Earth:

you can have one post-it note at a time

you can make one or two clicks to move between post-it notes.

…that’s not so good.


For the CRM systems that I have seen, that is the state of art! Some managers wonder why many sales people have a low regard for CRM systems: they think of them as a way of being controlled, rather than a way of helping them sell!

Welcome to the state of art of most CRM, ERP and groupware systems.

HTC Desire HD…please make my smart-ass phone into a smarter phone

It’s the most wonderful phone, a kind of electronics swiss army electronic gadgetry: sat nav, camera, camcorder, music player…

The trouble is that it is smart ass phone. You know the one: it talks the talk , but lets you down; a phone that needs charging several times in the day for you to be sure that you will be able to make and receive phone calls. The problem is power consumption. Using Juice Plotter, I can report that power consumption is about 25% battery used per hour on active use, but still 15-25% battery used per hour if you leave it lying around without shutting stuff down.

Various kind people have suggested some solutions: carry around a spare battery, a USB charger, or even manually switch off sat nav, bluetooth, wifi, GSM data, all apps (apart from the music player!) before putting the phone into sleep mode. Is that really the only option? What I think that would be most useful is an app that

  • shuts everything sat nav, bluetooth, wifi, GSM data, all apps (apart from the music player!) down
  • wakes up every 15 mins and kills anything that has snuck in again and started up
  • shuts things down when you go into sleep or after 15 mins inactivity
  • has really clear and obvious controls
  • does not do anything else

If it is clearly testable, and it’s clear what it is doing, I believe it will be popular with most Android users, who have been thoroughly “burnt” by the power problem. Having got to the point where I am considering selling my phone, despite its many excellent features, I would definitely be prepared to spend money on such an application.

I need a smart phonenot a smart-ass phone.