Planning Analytics

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Hi Paul,

I doubt you'll get an unbiased response here, so I won't pretend to give one. I'll highlight some off differences that those who have chosen Planning Analytics over Anaplan have identified as decision points.
1. Data Scale - the TM1 database that's at the core of Planning Analytics is difficult to beat in terms of data scale because it handles sparsity so well. It handles very high volumes and still maintains the ability to do instant recalculations.
2. Excel Add-in - I understand Anaplan has an Excel Add-in, but Planning Analytics two decade record of supporting Excel as a primary interface means that it's functionality is very robust.
3. On Prem Support - Both Anaplan an Planning Analytics support Saas deployments. Planning Analytics gives you the option of keeping your application on prem
4. Customizability - We usually hear that Anaplan is fast to deploy, but only if you take their out of the box solution. If your company does something differently then Planning Analytics has a long track record of very flexible solutions.
So these are the main positive differentiators for PA vs Anaplan. I'd encourage you to look at some third party research to get a fuller picture.

Paul Glennon

------------------------------
PAUL GLENNON
------------------------------

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Hi Paul Glennon

I hope all is well!

• Anaplan provides itself when it comes to pre-built templates/solutions. I started building pre-built templates that can be used by finance - Box

  Box		remove preview
  Box View this on Box >

• Anaplan focus has been on supply chain management. PAoC does a better job with supply chain management - Box

  Box		remove preview
  Box View this on Box >

• Anaplan has issues with memory management which PAoC does a better job with memory management.

------------------------------
paul YOUNG
------------------------------

Original Message:
Sent: Tue May 26, 2020 09:02 AM
From: PAUL GLENNON
Subject: IBM Planning Analytics vs Anaplan

Hi Paul,

I doubt you'll get an unbiased response here, so I won't pretend to give one. I'll highlight some off differences that those who have chosen Planning Analytics over Anaplan have identified as decision points.
1. Data Scale - the TM1 database that's at the core of Planning Analytics is difficult to beat in terms of data scale because it handles sparsity so well. It handles very high volumes and still maintains the ability to do instant recalculations.
2. Excel Add-in - I understand Anaplan has an Excel Add-in, but Planning Analytics two decade record of supporting Excel as a primary interface means that it's functionality is very robust.
3. On Prem Support - Both Anaplan an Planning Analytics support Saas deployments. Planning Analytics gives you the option of keeping your application on prem
4. Customizability - We usually hear that Anaplan is fast to deploy, but only if you take their out of the box solution. If your company does something differently then Planning Analytics has a long track record of very flexible solutions.
So these are the main positive differentiators for PA vs Anaplan. I'd encourage you to look at some third party research to get a fuller picture.

Paul Glennon

------------------------------
PAUL GLENNON

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Couple of other points

• Hierarchies - To my knowledge you cannot view multiple hierarchies from the same dimension within a view
• ETL Tool - While they do have a small bat file like scripting tool that can move data they almost always push for buying the informatica connector in turn adding to the overall cost.
• No visualization - Out of the box they don't support visualization.  They would tell you to buy Tableau in which they have a connector.  This further adds to cost and can cause data latency and sprawl as many users attempt to get data "Out" of Anaplan
• Data Scale - Last I checked they charge based on memory size.  Unfortunately since their technology uses memory where their is sparsity you end up paying for sparsity (Essentially empty cells).

------------------------------
Robby Meyers
------------------------------

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Paul Glenn:

For me, the Anaplan deal-breaker is its (lack of) sparsity management, which means it simply cannot scale. Suppose you have a sales model with 12 time periods, 20 products, and 10 stores. The theoretical size of this model is 12 x 20 x 10 = 2,400. If every store sells every product in every time period, you'll have 2,400 sales numbers. But if on average only half the products sell each month in each store, you'll just have about 1,200 sales numbers.

Anaplan will set aside storage for 2,400 numbers regardless of whether you actually have 2,400 numbers. IBM Planning Analytics (and TM1) will only set aside storage for 1,200. That's what sparsity management means.

This does not sound too bad if 2,400 numbers is as big as your model gets, but the problem comes with exponential growth. Imagine you have 200,000 products and 1,000 stores as some of our retail customers do. I can guarantee that not every product is sold in every store in every month, but Anaplan would need to set aside storage for 2.4 billion numbers, while IBM Planning Analytics (and TM1) would make do with just a fraction (and does). Conservatively, if a number is stored in 8 bytes, and you want the speed of an in-memory architecture, you would need a computer with 18,000 gigabytes for just this model!

Anaplan solves this problem by requiring that you break up your model into pieces, which introduces the new problem of getting a consolidated view. Furthermore, I am told that their pricing is tied directly to model size, which is a way of discouraging customers from building detailed models.

Those of us who used to work with the old Cognos Enterprise Planning product (formerly Adaytum) are very familiar with the pain of working around this kind of database explosion. 

If your models are small (closer to the 2,400 number example) and unlikely to grow, you will be able to make do with Anaplan. But for anything else beware of this mathematical limitation. 

AG Tan
QueBIT

------------------------------
Ann-Grete Tan
------------------------------

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Paul, As well as the differences outlined by Palu Glennon below I would add a few others:

• Planning Analytics (PA) has a much more powerful calculation engine for more complex processes. As well as standard budgeting systems we have developed industry specific applications for insurance premium modelling, production planning and aircraft ground handling costs.  All of which involve high volume sophisticated modelling
• PA has sophisticated, easy to use presentation tools for reporting, analysis and input. The Workspace provides full browser based dashboarding interface and you can use the Excel add-in to design higly formatted Excel templates for input and reporting.  These can be accessed either directly in Excel or as a "web sheet" in the Workspace.
• Infrastructure.  PA can be provided as a full SaaS offering or via a perpetual license that can be installed on premises and on a public or private cloud.  The software is exaclty the same across all platforms and you can easily move application between the platforms. 
• Pricing.  In general I think you will find PA significantty cheaper than Anaplan, particularly if you consider that all users are licensed to create their own reports and dashboards.  My understanding is that, with Anaplan, you need to be an adminsirator to create dashboards and many Anaplan customers need to purchase  Tableau to do the reporting.  Happy to be corrected if this is no the case. 

Hope this helps

------------------------------
stephen waters
------------------------------

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Hi

Obviously you are not likely to get an unbiased opinion here.

From what I have heard Anaplan pretty much runs out of steam at 5 dimensions, largely related to the sparsity issue. 

If you take a typical model, you have a minimum of :

1) Version dimension, Actual, Budget, Forecast, etc
2&3) At least one Time dimension, although for Forecasting you usually want two, eg how does the forecast for this year made this month compare with the forecast for this year made last month, etc.
3,4,5) Currency conversion, even a simple list of input currencies converting to a single company currency needs one dimension. If you have to convert to more than one currency, eg expressing figures in dollars and pounds, that's another dimension, and if you want to do things like being able to express data at historic rates or current rates to be able to eliminate or reflect the effect of currency exchange movements, that's another dimension
6) Measures dimension, eg Units, and Sales Value, etc
7) Probably some sort of adjustment dimension will be needed, eg figures before and after adjustments, figures before and after central expense allocations.

So you can easily use up 7 dimensions before you have put any business specific dimensions into the model, eg Entity, Department, Product, Geography, Project, etc.

How do Anaplan handle something like this? Well from what I have heard, you will probably find lots of little cubes, one for Actuals, one for Budget, etc, one for pounds, one for dollars, etc. You have to push data from one cube to another. It rapidly gets more complex to build. Pushing data from one cube to another means that it could hardly be described as real-time.

By comparison you can do this all in a single cube in TM1/PA and get far more insights in the data, because the modelling is better. If you change a number in the TM1 cube you can instantly see the effect of that change at all levels of consolidation, in all currencies, etc. Because things like Forecast and Budget are in the same Version dimension in the same cube, you can instantly get Forecast vs Budget variance. You cannot do that if they are held in separate little cubes in Anaplan. 

I would see Anaplan as a solution for small department with limited modelling requirements. It is unlikely to work as an enterprise solution.

Unfortunately, issues like scalability and the ability to handle complex modelling are often not apparent in the quick demos that most customers see before buying a product. 

Regards

Paul Simon



------------------------------
Paul Simon
------------------------------

Original Message:
Sent: Tue May 26, 2020 08:30 PM
From: stephen waters
Subject: IBM Planning Analytics vs Anaplan

Paul, As well as the differences outlined by Palu Glennon below I would add a few others:

• Planning Analytics (PA) has a much more powerful calculation engine for more complex processes. As well as standard budgeting systems we have developed industry specific applications for insurance premium modelling, production planning and aircraft ground handling costs.  All of which involve high volume sophisticated modelling
• PA has sophisticated, easy to use presentation tools for reporting, analysis and input. The Workspace provides full browser based dashboarding interface and you can use the Excel add-in to design higly formatted Excel templates for input and reporting.  These can be accessed either directly in Excel or as a "web sheet" in the Workspace.
• Infrastructure.  PA can be provided as a full SaaS offering or via a perpetual license that can be installed on premises and on a public or private cloud.  The software is exaclty the same across all platforms and you can easily move application between the platforms. 
• Pricing.  In general I think you will find PA significantty cheaper than Anaplan, particularly if you consider that all users are licensed to create their own reports and dashboards.  My understanding is that, with Anaplan, you need to be an adminsirator to create dashboards and many Anaplan customers need to purchase  Tableau to do the reporting.  Happy to be corrected if this is no the case. 

Hope this helps

------------------------------
stephen waters

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Hi,

Analplan is good for prototyping or a very fluid model, because you can add or remove dimensions from cubes, and the interface is easier for new users.

Anaplan only allows calculations along the measure dimension of a cube.  TM1/PA has the ability to tailor calculations on multiple intersections or even for a specific cell.


------------------------------
Scott Brown
------------------------------

Original Message:
Sent: Wed May 27, 2020 02:57 PM
From: Paul Simon
Subject: IBM Planning Analytics vs Anaplan

Hi

Obviously you are not likely to get an unbiased opinion here.

From what I have heard Anaplan pretty much runs out of steam at 5 dimensions, largely related to the sparsity issue. 

If you take a typical model, you have a minimum of :

1) Version dimension, Actual, Budget, Forecast, etc
2&3) At least one Time dimension, although for Forecasting you usually want two, eg how does the forecast for this year made this month compare with the forecast for this year made last month, etc.
3,4,5) Currency conversion, even a simple list of input currencies converting to a single company currency needs one dimension. If you have to convert to more than one currency, eg expressing figures in dollars and pounds, that's another dimension, and if you want to do things like being able to express data at historic rates or current rates to be able to eliminate or reflect the effect of currency exchange movements, that's another dimension
6) Measures dimension, eg Units, and Sales Value, etc
7) Probably some sort of adjustment dimension will be needed, eg figures before and after adjustments, figures before and after central expense allocations.

So you can easily use up 7 dimensions before you have put any business specific dimensions into the model, eg Entity, Department, Product, Geography, Project, etc.

How do Anaplan handle something like this? Well from what I have heard, you will probably find lots of little cubes, one for Actuals, one for Budget, etc, one for pounds, one for dollars, etc. You have to push data from one cube to another. It rapidly gets more complex to build. Pushing data from one cube to another means that it could hardly be described as real-time.

By comparison you can do this all in a single cube in TM1/PA and get far more insights in the data, because the modelling is better. If you change a number in the TM1 cube you can instantly see the effect of that change at all levels of consolidation, in all currencies, etc. Because things like Forecast and Budget are in the same Version dimension in the same cube, you can instantly get Forecast vs Budget variance. You cannot do that if they are held in separate little cubes in Anaplan. 

I would see Anaplan as a solution for small department with limited modelling requirements. It is unlikely to work as an enterprise solution.

Unfortunately, issues like scalability and the ability to handle complex modelling are often not apparent in the quick demos that most customers see before buying a product. 

Regards

Paul Simon



------------------------------
Paul Simon

Original Message:
Sent: Tue May 26, 2020 08:30 PM
From: stephen waters
Subject: IBM Planning Analytics vs Anaplan

Paul, As well as the differences outlined by Palu Glennon below I would add a few others:

• Planning Analytics (PA) has a much more powerful calculation engine for more complex processes. As well as standard budgeting systems we have developed industry specific applications for insurance premium modelling, production planning and aircraft ground handling costs.  All of which involve high volume sophisticated modelling
• PA has sophisticated, easy to use presentation tools for reporting, analysis and input. The Workspace provides full browser based dashboarding interface and you can use the Excel add-in to design higly formatted Excel templates for input and reporting.  These can be accessed either directly in Excel or as a "web sheet" in the Workspace.
• Infrastructure.  PA can be provided as a full SaaS offering or via a perpetual license that can be installed on premises and on a public or private cloud.  The software is exaclty the same across all platforms and you can easily move application between the platforms. 
• Pricing.  In general I think you will find PA significantty cheaper than Anaplan, particularly if you consider that all users are licensed to create their own reports and dashboards.  My understanding is that, with Anaplan, you need to be an adminsirator to create dashboards and many Anaplan customers need to purchase  Tableau to do the reporting.  Happy to be corrected if this is no the case. 

Hope this helps

------------------------------
stephen waters

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson

Loving the typo!  Juvenile I know...
I've not used Anaplan but as a developer in PA its straight forward to rebuild cubes when in the initial prototyping stage of an appliction and post this we have Hierarchies to deliver additional analysis dimensions as requirements change.

------------------------------
Steven Rowe
------------------------------

Original Message:
Sent: Mon June 01, 2020 12:20 PM
From: Scott Brown
Subject: IBM Planning Analytics vs Anaplan

Hi,

Analplan is good for prototyping or a very fluid model, because you can add or remove dimensions from cubes, and the interface is easier for new users.

Anaplan only allows calculations along the measure dimension of a cube.  TM1/PA has the ability to tailor calculations on multiple intersections or even for a specific cell.


------------------------------
Scott Brown

Original Message:
Sent: Wed May 27, 2020 02:57 PM
From: Paul Simon
Subject: IBM Planning Analytics vs Anaplan

Hi

Obviously you are not likely to get an unbiased opinion here.

From what I have heard Anaplan pretty much runs out of steam at 5 dimensions, largely related to the sparsity issue. 

If you take a typical model, you have a minimum of :

1) Version dimension, Actual, Budget, Forecast, etc
2&3) At least one Time dimension, although for Forecasting you usually want two, eg how does the forecast for this year made this month compare with the forecast for this year made last month, etc.
3,4,5) Currency conversion, even a simple list of input currencies converting to a single company currency needs one dimension. If you have to convert to more than one currency, eg expressing figures in dollars and pounds, that's another dimension, and if you want to do things like being able to express data at historic rates or current rates to be able to eliminate or reflect the effect of currency exchange movements, that's another dimension
6) Measures dimension, eg Units, and Sales Value, etc
7) Probably some sort of adjustment dimension will be needed, eg figures before and after adjustments, figures before and after central expense allocations.

So you can easily use up 7 dimensions before you have put any business specific dimensions into the model, eg Entity, Department, Product, Geography, Project, etc.

How do Anaplan handle something like this? Well from what I have heard, you will probably find lots of little cubes, one for Actuals, one for Budget, etc, one for pounds, one for dollars, etc. You have to push data from one cube to another. It rapidly gets more complex to build. Pushing data from one cube to another means that it could hardly be described as real-time.

By comparison you can do this all in a single cube in TM1/PA and get far more insights in the data, because the modelling is better. If you change a number in the TM1 cube you can instantly see the effect of that change at all levels of consolidation, in all currencies, etc. Because things like Forecast and Budget are in the same Version dimension in the same cube, you can instantly get Forecast vs Budget variance. You cannot do that if they are held in separate little cubes in Anaplan. 

I would see Anaplan as a solution for small department with limited modelling requirements. It is unlikely to work as an enterprise solution.

Unfortunately, issues like scalability and the ability to handle complex modelling are often not apparent in the quick demos that most customers see before buying a product. 

Regards

Paul Simon



------------------------------
Paul Simon

Original Message:
Sent: Tue May 26, 2020 08:30 PM
From: stephen waters
Subject: IBM Planning Analytics vs Anaplan

Paul, As well as the differences outlined by Palu Glennon below I would add a few others:

• Planning Analytics (PA) has a much more powerful calculation engine for more complex processes. As well as standard budgeting systems we have developed industry specific applications for insurance premium modelling, production planning and aircraft ground handling costs.  All of which involve high volume sophisticated modelling
• PA has sophisticated, easy to use presentation tools for reporting, analysis and input. The Workspace provides full browser based dashboarding interface and you can use the Excel add-in to design higly formatted Excel templates for input and reporting.  These can be accessed either directly in Excel or as a "web sheet" in the Workspace.
• Infrastructure.  PA can be provided as a full SaaS offering or via a perpetual license that can be installed on premises and on a public or private cloud.  The software is exaclty the same across all platforms and you can easily move application between the platforms. 
• Pricing.  In general I think you will find PA significantty cheaper than Anaplan, particularly if you consider that all users are licensed to create their own reports and dashboards.  My understanding is that, with Anaplan, you need to be an adminsirator to create dashboards and many Anaplan customers need to purchase  Tableau to do the reporting.  Happy to be corrected if this is no the case. 

Hope this helps

------------------------------
stephen waters

Original Message:
Sent: Sun May 24, 2020 10:10 AM
From: Paul Glenn
Subject: IBM Planning Analytics vs Anaplan

What are the differences (technical limitations) of IBM Planning Analytics (not old TM1) vs Anaplan?

------------------------------
Paul Glenn
------------------------------

#PlanningAnalyticswithWatson