Many CEOs try to stay out of Salesforce implementation, but this can be a big mistake. Learn how CEOs can set a Salesforce implementation up for success. Sales Analytics Solution. What CEOs Must Know About Salesforce Implementation. By Mike Baker. Posted September 4, 2015. Jun 20, 2013 With that in mind, Forrester used the following screening criteria to pick which companies it considers in its latest Wave ranking for Salesforce.com implementation partners. They include.
If you’ve seen Fight Club, you probably remember the scene when Brad Pitt’s character says, “You are not special. You’re not a beautiful and unique snowflake.” As hard as that may be for many of us to hear, let alone believe, it’s very relevant for companies when it comes to their analytics.When it comes to our businesses, we overestimate our singularity. Many companies think that the only analytics solution that will work for them is a custom one – their business being so unique, so like that snowflake, that no preconfigured platform or tool could truly be applicable to what they do. This thought is often subconsciously or directly promoted by data scientists or consultants who want to do cool stuff that breaks new ground or builds new skills.
Of course, every company has unique problems to solve. The key skill is choosing to solve the problems that have the maximum business impact. From what I have seen in the field, too many companies first focus on hitting a grand slam through a custom solution rather than learning how to hit 100 singles with a well-understood, but perhaps less glamorous application of analytics.
This article promotes a simple recommendation: To achieve maximum impact, apply as many productized analytic solutions as you can rather than attempting to solve unique challenges through custom analytics.
A refresher - What are productized analytics?
Earlier this year, I wrote a case study of productized analytics about ClickFox, in which I explained how productized analytics are a way to use a pre-designed analytics tool to solve a well-defined and understood business process. In order for a productized analytics to exist, it must be a common problem. In the case of ClickFox the common problem is specifically customer journey analytics, which can be seen as a special case of journey analytics.
But if we look at what Teradata, IBM, Salesforce, Adobe and SAP are bringing to market, we can see the race is on to provide advanced analytics in the context of a well understood business process.
Productized analytics allow companies to buy an out of the box solution for a common problem. There’s no need to redesign the wheel in every analytics instance after all. Like buying a car off the lot, rather than building a special purpose vehicle, productized analytics will get most companies where they need to go even if the way it is done isn’t particularly flashy.
My argument is that the risk of using a productized analytic is lower than attempting to invent a new solution for a unique problem. I’m willing to admit right off the bat that productized analytics in a narrow sense will not create as much competitive differentiation as a successful custom analytic.
But, I assert that using as many productized analytics as possible can allow a company to rapidly expand the impact of data on many aspects of the business for the following reasons:
- Accelerating adoption of analytics will be differentiating.
- Wide application of analytics will increase the sophistication of the staff both about the impact of data and about the value of analytics and the fit of various types of analytics to various problems. This will expand the number of high quality ideas for custom analytics.
- Because most productized analytics have lots of degrees of freedom and can be customized. Once the vanilla version is implemented and working well, the analytics can be customized to be more effective by adding more data, better data, or improving the tuning of the analytics.
What is the definition of a Productized Analytic?
My definition of a productized analytic covers a broad amount of technology coming to market. Salesforce’s Einstein, Adobe Analytics, Google Analytics, Teradata’s collection Analytics Solutions implemented through Rapid Analytics Consulting Engagements all are forms of productized analytics. The bulk of the capabilities of ClickFox and Qubit and many products in the marketing space are productized analytics.
Here is my working definition. A productized analytic is a solution that has a clear opinion and suggested implementation in most of these dimensions:
- Sources of data
- An ingestion process
- Structure of a generic data platform
- Analytics or application specific data objects
- Data model and analytics
- Reports and visualizations
- Data exports and API
- Actions that control or influence a business process
- Feedback mechanisms to support continuous improvement
Not every dimension needs to be defined to make up a productized analytic. The last two for example, taking action and getting feed back, are not present in many solutions.
Why do productized analytics make sense?
Many companies have a view that the only way to achieve game-changing analytics is through a superstar data scientist. This doesn’t make sense for many reasons. For one, there aren’t enough superstars out there. And the ones that do exist cost a premium. Even if you are one of the lucky companies out there to find these unicorns, there’s so much competition for their services that you will either have to constantly up their compensation or just accept constant churn at these positions. Neither way is a sustainable approach. As we’re in the midst of football season, I’ll turn to a sports metaphor – too many franchises have been lured by the big name free agent signing as a supposed cure-all to make them a championship contender. In reality, across all sports, teams still win, and if you invest too heavily in one or two players, it negatively impacts the team as a whole.
Additionally, while the custom solution a rock star data scientist could provide you might meet each of your needs, all of the analytic prowess of your organization will be funneled through a single individual or department. That inevitably leads to processing lags and overdependence on people with specialized skill sets.
Productized analytics, on the other hand, democratize analytics, bringing them to the masses. The tools are broadly targeted, usually straightforward and simple to use, meaning even the non-tech savvy can suddenly rely on data and analytics to better understand their roles and their customers. And this is what’s so exciting to me about the potential of productized analytics – just as the Gutenberg Press opened up literacy to the world with astounding results, bringing analytics to more people means data can become the lifeblood of the entire company, rather than just a select few. So salespeople can get the answers they want immediately, rather than waiting for weeks when requests back up. And once they get those answers, they’ll have more questions, furthering the ability of data to change the business for the better.
What’s crucial here to remember is that using productized analytics doesn’t confine a company in the long-term. Why is this? Well, I think it’s helpful to think of productized analytics not as a frozen application, but more like an analytics starter kit. They’re a way to get more people in your company excited about using analytics and data to drive their decisions. Once the basic analytics have become operational, companies can differentiate and customized to their individual business needs, or by incorporating new data sources. And just as importantly, how each company reacts to the information unleashed by the analytics will be highly unique. It could lead to greater investments in analytics for some companies, or just new questions people want answered in others.
That’s why I look at productized analytics as a form of relevant innovation. A few years ago, Atul Gawande, a renowned surgeon and author, wrote a piece for The New Yorker, about the need for health care to become less like the French Laundry, and more like the Cheesecake Factory. Namely, he argued that health care had far too little standardization and that innovation in the field wouldn’t be more specialized techniques, but rather more uniformity across hospitals and providers in using the techniques already proven to work.
In many ways, analytics are similar. Yes, there are some cases where a truly unique solution is needed. But for most companies, the goal should be to ingrain and implement an analytics model as quickly as possible. If some customization is lost in the short-term because of that, it’s a trade-off worth making. After all, the democratization and liberalization of analytics that productized analytics offer are what could truly provide a business with the most unique thing of all – sustained success.
Follow Dan Woods on Twitter:
Follow @danwoodsearly
Dan Woods is on a mission to help people find the technology they need to succeed. Users of technology should visit CITO Research, a publication where early adopters find technology that matters. Vendors should visit Evolved Media for advice about how to find the right buyers. See list of Dan's clients on this page.
'>If you’ve seen Fight Club, you probably remember the scene when Brad Pitt’s character says, “You are not special. You’re not a beautiful and unique snowflake.” As hard as that may be for many of us to hear, let alone believe, it’s very relevant for companies when it comes to their analytics.
When it comes to our businesses, we overestimate our singularity. Many companies think that the only analytics solution that will work for them is a custom one – their business being so unique, so like that snowflake, that no preconfigured platform or tool could truly be applicable to what they do. This thought is often subconsciously or directly promoted by data scientists or consultants who want to do cool stuff that breaks new ground or builds new skills.
Don't Swing for the Fences, Instead Hit Singles with Productized Analytics/Chris Denorfia of Major League Baseball's (MLB) Oakland Athletics hits a run-scoring single against the Boston Red Sox during the second inning of Game 2 of the Richo Japan Opening Series 2008 at the Tokyo Dome in Tokyo, Japan, on Wednesday, March 26, 2008. Pitcher Rich Harden allowed a run on three hits and Emil Brown hit a three-run home run to lead the Oakland Athletics over the Boston Red Sox 5-1 and earn a split of the two-game season-opening series in Japan. Photographer: Haruyoshi Yamaguchi/Bloomberg News
Of course, every company has unique problems to solve. The key skill is choosing to solve the problems that have the maximum business impact. From what I have seen in the field, too many companies first focus on hitting a grand slam through a custom solution rather than learning how to hit 100 singles with a well-understood, but perhaps less glamorous application of analytics.
This article promotes a simple recommendation: To achieve maximum impact, apply as many productized analytic solutions as you can rather than attempting to solve unique challenges through custom analytics.
A refresher - What are productized analytics?
Earlier this year, I wrote a case study of productized analytics about ClickFox, in which I explained how productized analytics are a way to use a pre-designed analytics tool to solve a well-defined and understood business process. In order for a productized analytics to exist, it must be a common problem. In the case of ClickFox the common problem is specifically customer journey analytics, which can be seen as a special case of journey analytics.
But if we look at what Teradata, IBM, Salesforce, Adobe and SAP are bringing to market, we can see the race is on to provide advanced analytics in the context of a well understood business process.
Productized analytics allow companies to buy an out of the box solution for a common problem. There’s no need to redesign the wheel in every analytics instance after all. Like buying a car off the lot, rather than building a special purpose vehicle, productized analytics will get most companies where they need to go even if the way it is done isn’t particularly flashy.
My argument is that the risk of using a productized analytic is lower than attempting to invent a new solution for a unique problem. I’m willing to admit right off the bat that productized analytics in a narrow sense will not create as much competitive differentiation as a successful custom analytic.
But, I assert that using as many productized analytics as possible can allow a company to rapidly expand the impact of data on many aspects of the business for the following reasons:
- Accelerating adoption of analytics will be differentiating.
- Wide application of analytics will increase the sophistication of the staff both about the impact of data and about the value of analytics and the fit of various types of analytics to various problems. This will expand the number of high quality ideas for custom analytics.
- Because most productized analytics have lots of degrees of freedom and can be customized. Once the vanilla version is implemented and working well, the analytics can be customized to be more effective by adding more data, better data, or improving the tuning of the analytics.
What is the definition of a Productized Analytic?
My definition of a productized analytic covers a broad amount of technology coming to market. Salesforce’s Einstein, Adobe Analytics, Google Analytics, Teradata’s collection Analytics Solutions implemented through Rapid Analytics Consulting Engagements all are forms of productized analytics. The bulk of the capabilities of ClickFox and Qubit and many products in the marketing space are productized analytics.
Here is my working definition. A productized analytic is a solution that has a clear opinion and suggested implementation in most of these dimensions:
- Sources of data
- An ingestion process
- Structure of a generic data platform
- Analytics or application specific data objects
- Data model and analytics
- Reports and visualizations
- Data exports and API
- Actions that control or influence a business process
- Feedback mechanisms to support continuous improvement
Not every dimension needs to be defined to make up a productized analytic. The last two for example, taking action and getting feed back, are not present in many solutions.
Why do productized analytics make sense?
Many companies have a view that the only way to achieve game-changing analytics is through a superstar data scientist. This doesn’t make sense for many reasons. For one, there aren’t enough superstars out there. And the ones that do exist cost a premium. Even if you are one of the lucky companies out there to find these unicorns, there’s so much competition for their services that you will either have to constantly up their compensation or just accept constant churn at these positions. Neither way is a sustainable approach. As we’re in the midst of football season, I’ll turn to a sports metaphor – too many franchises have been lured by the big name free agent signing as a supposed cure-all to make them a championship contender. In reality, across all sports, teams still win, and if you invest too heavily in one or two players, it negatively impacts the team as a whole.
Additionally, while the custom solution a rock star data scientist could provide you might meet each of your needs, all of the analytic prowess of your organization will be funneled through a single individual or department. That inevitably leads to processing lags and overdependence on people with specialized skill sets.
Productized analytics, on the other hand, democratize analytics, bringing them to the masses. The tools are broadly targeted, usually straightforward and simple to use, meaning even the non-tech savvy can suddenly rely on data and analytics to better understand their roles and their customers. And this is what’s so exciting to me about the potential of productized analytics – just as the Gutenberg Press opened up literacy to the world with astounding results, bringing analytics to more people means data can become the lifeblood of the entire company, rather than just a select few. So salespeople can get the answers they want immediately, rather than waiting for weeks when requests back up. And once they get those answers, they’ll have more questions, furthering the ability of data to change the business for the better.
What’s crucial here to remember is that using productized analytics doesn’t confine a company in the long-term. Why is this? Well, I think it’s helpful to think of productized analytics not as a frozen application, but more like an analytics starter kit. They’re a way to get more people in your company excited about using analytics and data to drive their decisions. Once the basic analytics have become operational, companies can differentiate and customized to their individual business needs, or by incorporating new data sources. And just as importantly, how each company reacts to the information unleashed by the analytics will be highly unique. It could lead to greater investments in analytics for some companies, or just new questions people want answered in others.
That’s why I look at productized analytics as a form of relevant innovation. A few years ago, Atul Gawande, a renowned surgeon and author, wrote a piece for The New Yorker, about the need for health care to become less like the French Laundry, and more like the Cheesecake Factory. Namely, he argued that health care had far too little standardization and that innovation in the field wouldn’t be more specialized techniques, but rather more uniformity across hospitals and providers in using the techniques already proven to work.
In many ways, analytics are similar. Yes, there are some cases where a truly unique solution is needed. But for most companies, the goal should be to ingrain and implement an analytics model as quickly as possible. If some customization is lost in the short-term because of that, it’s a trade-off worth making. After all, the democratization and liberalization of analytics that productized analytics offer are what could truly provide a business with the most unique thing of all – sustained success.
Follow Dan Woods on Twitter:
Dan Woods is on a mission to help people find the technology they need to succeed. Users of technology should visit CITO Research, a publication where early adopters find technology that matters. Vendors should visit Evolved Media for advice about how to find the right buyers. See list of Dan's clients on this page.
As part of my continuing series on productized analytics and the potential benefits they can bring to companies, I recently spoke with Ketan Karkhanis, SVP & GM of Analytics Cloud at Salesforce. I wanted to understand what Salesforce offers in the productized analytics, or apps, space and the approach the company takes to those products.As I’ve discussed in the previous articles in this series, many companies still assume that the only way they can get an analytics solution that meets their needs is through a customized solution. In truth, there are many products now on the market that can meet the majority of analytic needs of most companies, while also including features that allow for some customization. The most significant benefit of using these productized analytics, or apps, is that they help to launch a company on their analytics journey, putting analytics and data in the hands of more people within the business to help guide decisions. This speeds up the process of discovery compared to relying on data scientists or a fully customized solution.
Ketan Karkhanis, SVP & GM of Analytics Cloud at Salesforce
I view productized analytics as existing in four tiers of offerings, as outlined in this infographic from Evolved Media.
- Stage Four: The highest stage of productization is the Value Meal, or one size fits all analytic, like a FitBit, that allows you to begin working with data in a consistent way but without any individualized bells or whistles.
- Stage Three: The third stage is the artisanal brew, like the soy mocha latte you get at Starbucks. There’s a fairly large menu of offerings in this set of products that allows users to get a greater degree of customization but still does not require them to create the analytics on their own.
- Stage Two: The second stage is the dinner in a box, or Blue Apron tier. With this level of product, there’s much more customization and flexibility possible, but users are given an analytics platform with all the ingredients to make that customization possible. But more creation on their part is necessary.
- Stage One: And then the lowest stage of productization is the custom kitchen, or French Laundry tier, which allows users to have full control over the design and creation of their analytics approach. This is where you get highly customized analytic platforms like Teradata’s Aster, Qlik's QlikSense, TIBCO Spotfire, Tableau, Sisense, and so on, which support creation of solutions custom designed for the specific needs of a customer.
My argument has been that most companies think they need the custom kitchen, when the artisanal brew or dinner in a box will suffice – at least to get started.I spoke with Karkhanis about where Salesforce’s products fit into this paradigm, as well as the possibilities those products open up for companies.
Salesforce’s definition of a productized analytic
It was important to hear Salesforce’s definition of a productized analytic tool in order to know where the company’s products fit in as compared to what else is on the market. To get to that definition, Karkhanis told me there are four basic questions any user wants to get out of using an analytic tool:
- What happened?
- Why did it happen?
- What will happen?
- What should I do about it?
According to Karkhanis, the company’s products are guided by a philosophy that analytics should be advanced, but also simple to use, and that they should touch upon business processes to actually affect change. A productized analytics product should be able to answer these questions without having to rely on the building of a complicated analytics stack or data science.
As I’ve outlined before, to answer these questions, a productized analytics tool can have a range of capabilities that can include the ability to ingest data from a variety of sources, structure a generic data platform, apply analytics to specific data objects that results in data models and analytics, and then have these findings available for exports and in reports, APIs, and visualizations. And the tools should help to inform businesses processes and have some ability to receive feedback to improve the analytic performance over time.
Karkhanis views Salesforce’s Analytics Cloud as a productized analytic tool because it can perform all these functions and be linked to a business process. Karkhanis believes it’s this last piece that is particularly important for any productized analytics product to be successful.
“What’s very important in the lens of an application for us is not just all the visualizations and actionability,” he said. “It is the shaping of behaviors. Because the app also espouses a point of view. And helps to guide future behavior.”
Where Salesforce’s products fit in
Karkhanis also told me that within Salesforce’s portfolio, their products mainly fit into the first, second, and third stages of the paradigm I outlined above. Here are some of the productized analytics tools they offer:
- Salesforce Analytics Cloud is their cloud analytics tool. It’s fully customizable, but truly designed to promote data and analytic democracy across an organization by being accessible and easy to use.
“Analytics Cloud allows users to arrive at an answer quicker than with other analytics products,” Karkhanis said. “Our goal with it is to empower every business user with the insights they need to make decisions, without needing complex data science.” He said that Analytics Cloud runs sales, service, IT and marketing analytics for business stakeholders. But he was also careful to emphasize that Analytics Cloud embeds in business processes because it is designed for all users to be able to leverage data to be “actionable.”
“The key for us is that Analytics Cloud combines business process, analytics, and engagement, in one tool, which is what a productized analytic product should do,” he said.
- BeyondCore is Salesforce’s smart data discovery tool, a product that is machine-led, and machine-learning based analytics designed, in the words of Karkhanis, to “eliminate bias in analyzing your data and allow you to come up with explanations of what is happening in your data much faster.”
“BeyondCore generates models on the fly and does advanced data science,” Karkhanis said. “But it’s got an inbuilt feedback mechanism because it is not machine-led only – there’s room for human-augmentation to improve the models.” Essentially, the tool allows for user feedback to improve results and the machine learning.
This level of customization places these products in the lower and middle tiers of productized analytics products I mentioned before – they can be dinner in a box, where you have all the ingredients to create a successful tool, but you can also create something on your own from scratch.
The Source of Product Insight and Analytical Approach
Bolstering this assessment is the approach Salesforce takes overall to their productized analytics products. That approach is predicated heavily on incorporating and implementing user feedback to improve product performance over time.According to Karkhanis, every stage of development of the Salesforce products is refined by user feedback.
“We have roughly 150,000 customers who can all provide feedback to help us improve our tools,” he said. “The way we start building an analytics app is to focus on the persona of our user, our customers, and we do a ton of research to understand what behaviors those people need, and their goals and outcomes. It’s a big listening loop with our customers – everyday we are getting feedback on what our customers are doing and how they are adopting our product. We design and improve our products based on those experiences.”
As a result of this approach, Salesforce can offer customers at different levels of analytic sophistication a range of tools to meet their needs. For some small businesses just starting on implementing analytics, Analytics Cloud can be used as a more off-the-shelf product. But customization can occur over time as their needs change and grow.
The Level of Productization and the Target Audience
Keeping the customer in mind enables Salesforce to have a range of offerings open to evolution and adaption. Karkhanis emphasized that their customers exist on a continuum, starting with users who want a Starbucks level stacks-tool, to those who become familiar with the tools and add complexity over time. Salesforce specifically targets business consumers and analysts, as well as data scientists with their products. But Karkhanis said that “When we design our products, we are not just looking at analysts, but at the entire line of business and who can achieve the greatest impact from using analytics.”
The Level of Process Integration
I’ve written before that data and analytics by themselves don’t drive insights or change. They must be integrated into business processes to lead to results. Karkhanis agreed, and repeatedly stressed the importance of embedding analytics in business processes during our conversation. He said the interface for the Analytics Cloud is designed to be as integrated as possible – so for a service analytics client, customer data is analyzed in the customer service platform.
“Bringing analytics and process together is a big element of our strategy,” he said. “Analytics are embedded in every part of your sales business process [with Analytics Cloud]. For sales teams, for instance, every opportunity is linked to an account that shows the account’s lifetime value, what products have or haven’t been sold to the account, and which have been sold to similar accounts.”
He mentioned that Salesforce also has its new AI offering across all clouds called Einstein, which creates a predictive score based on that history for sales members to be able to better target their customers.
The Implementation Platform
In reviewing the range of offerings on the market, it’s clear that some productized analytics are primarily intellectual property that is completely divorced from any implementation strategy. Others are only offered in the context of a specific product or application. Understanding how tightly coupled a productized analytic is to a specific application, implementation, or platform is crucial to understand when it is appropriate to use.
Karkhanis agreed with this assessment and said that it helps to guide Salesforce development process. “Our analytics platform is part of the Salesforce customer success platform,” he said. “On top of the analytics layer are productized packaged applications for sales, service, and marketing so customers get a seamless integration. This makes it incredibly easy to use.”
The goal with any productized analytic product is to have it be easily incorporated into the way staff already and will use data in their work. It can’t be an additional burden or people will not use it.
The Sweet Spot
Obviously, the goal of any productized analytics adoption is to get results and understand that sweet spot of out of the box tool and customized solution. But, how do you know when a productized analytic a good fit?
For Karkhanis, the answer to that questions lies in another question, namely “Why did something happen?” He believes Salesforce’s products allow customers to answer this vital question, which then leads to analytic discovery.
“With our analytic products, you are not looking at analytics once a month, but rather as every day as part of your job,” he said. “Traditional BI has always been about reflection. But our tools allow you to place data in context and act. It’s about being able to decide what to do. For us, it’s about spending less time away from your customers and business process and actually making decisions. You can’t expect a sales rep to open a 52-page monthly metric pack from some analyst to figure out how to approach a particular client. The sales rep needs that information in context while he’s making the call on his phone in real time.”
Karkhanis shared an insight I’ve made as well – that productized analytics are opening up the world of data in all new ways and helping companies to make better decisions. Because productized analytics allow companies to accelerate their analytics process, data can become more central to everyone’s work.
“Our focus with analytical applications is that we believe analytics have to play a fundamental role in shaping winning behaviors in every organization,” he said. Productized analytic tools, he continued, “are not only about avoiding the cost and time of developing analytics, but also avoiding the work you have to do to figure out what’s important and what’s not.”