73criz

How to Assess Your Data Requirements When Implementing a Customer Data Platform

November 23, 2018 10:57 am No Comments November 23, 2018

This article is part of our CDP series. To read other articles in this series, click here.

Customer data is no longer just an analytics requirement, or something used to send out marketing messages – it is the source of critical competitive advantage for a B2C business.

Truly customer-centric organizations use customer data for all aspect of business – from guiding product roadmap and categories, designing rich experiences, pricing their offerings right, creating the right customer service interfaces, deciding sales channels and even to make the right hiring decisions. Modern marketers use these customer insights to form a unified understanding of every customer and create contextual experiences that strengthen the relationship with the customer, ensuring loyalty and higher word of mouth.

Every company today collects more data about their customers that they use. They can also easily acquire additional, reliable data from data companies, and each byte presents insights that were not available before.

As a first step when deploying a Customer Data Platform (CDP), know what use cases you want to enable – this decides what data needs to be brought together. At this stage, assess your present data and what you need to acquire. Controlling customer churn, lifecycle marketing, and marketing attribution all require certain unique data sets.

The use cases and user types will dictate the format to store data in, the frequency of refresh and ingestion (real-time or batch) data, transformation required and responsiveness.

For example, in an e-commerce retail business like Amazon, purchase history is needed to provide personalized recommendations to customers. While digital data such as web-browsing/ app-browsing, items added to cart, saved for later, searched or viewed is captured, there is value in plugging in additional information such as income bracket, size of family, lifestyle, nature of work, types of vacations taken. This provides deep insights into what the customer would like, and the customer is pleasantly surprised with the retailer’s next-best product recommendations.

Organization owned customer data from the CDP foundation

Let us start with the data you have. If you have a Customer Relationship Management (CRM) program, provide any level of customer service or have a functioning loyalty program, then you have a solid foundation for your customer data platform. The CRM and loyalty systems can provide insights about demographics, purchase behavior, loyalty, and customer lifetime value and also contain the basic identity details.

But to do this right, it’s necessary to have a single customer record across all touchpoints by removing duplicate entries and resolving mismatches, that are created due to multiple reasons: incorrect data entry, different name variations of a customer, customers with multiple loyalty cards, change in their address etc. (Read more about solving the customer identity puzzle here). After de-duplication and linking customer records, you know the real size of your customer base and have more accurate profile snapshots.

Transactional and digital data generates actionable insights for strategic gains

If your business engages your customers in more than one channel, for example through the store and e-commerce, or over mobile apps, that’s a new set of data streaming in, which can enhance your understanding of your customers. Again, it’s important to link customer IDs across different channels, functions and customer devices to have a holistic customer view, that does not vary based on who is accessing it and where.

Most businesses use outbound digital communication data for pointed interventions, but information such as search, browsing patterns, offer engagement, wish-listing, abandoned carts all tell you something about the customer.

An insight-driven organization focuses on this knowledge, rather than the data itself. If along with the CRM and loyalty data, you have these astute observations, it can power better marketing and experience strategies. Storing every click of every customer is not feasible and honestly, not required. The signals you get by analyzing these clicks, cookies, customer location, and device information during the digital engagement should be part of your CDP strategy.

External data sets to enrich customer attributes

Finally, bring in 2nd and 3rd party web, demographic, psychographic, location, buying intent data to further enrich what you know about your customers. This is the ‘test-use-evaluate and iterates’ part of your data strategy. Not all data providers are the same, they all start their data journey differently and may or may not work for your brand and your customer base. It’s also important to note that data degrades quickly – the data was good a few quarters ago might not be valid anymore. Customer interests and intents may have evolved – they may have changed jobs or had a major life-event. To find out what works for you, do small experiments with a good set of data partners and choose the one that works best for your current marketing and strategy needs.

If you are poorly utilizing the data you have captured, or not capturing certain essential data, you are losing important revenue opportunities. If data is the new oil, accurate customer insights are high octane racing fuel for your B2C business!

Download CDP Handbook Now!

Retail Prophet Doug Stephens Interview: The store of the future won’t be a “store”

November 12, 2018 12:56 pm No Comments September 24, 2024

Changes are rampant in the retail industry. Every day we hear of both big brands (and small ones) closing shop or downsizing in an effort to stay lean and competitive. To understand in depth what’s happening in the retail landscape, we spoke to Doug Stephens.

Doug is the founder of Retail Prophet, and one of the world’s foremost retail industry futurists. The author of two groundbreaking books on retail, a nationally syndicated retail columnist, and over 20 years of experience in the retail industry, Doug brings together his unique perspective to provide Manthan with his insights on retailing, technology and consumer behavior.

MANTHAN: You speak of a sea change in retail and the trouble that’s coming for brick and mortar stores – can you explain this revolution?

DOUG: We have entered an era where media (in all its various form factors) has become “the store”. Whether it’s a magazine ad I can activate with my smartphone, my connected appliances, my smart TV or my in-home artificial intelligence, media is no longer a call-out to visit a store – it is the store. And this fact will only continue to escalate and permeate more aspects of our consumption. Therefore, the role and nature of physical stores must change in order to adapt.

Stores have to become places that move beyond the mere distribution of products and become enchanted spaces that distribute remarkable experiences. The problem is, most retailers aren’t internalizing this reality yet. Most store experiences remain unremarkable and it’s a fact that’s taking a toll on retail generally. If we’re honest, the retail brands that are getting wiped out aren’t being mourned because they had in fact become irrelevant long before they died.

“The retail brands that are getting wiped out aren’t being mourned because they had in fact become irrelevant long before they died.”

MANTHAN: What’s the role of technology in this revolution and which retailers are leading?

DOUG: Technology has essentially become the connective tissue in every aspect of our lives, whether we’re talking about business, leisure, commerce etc. In retail, we know that upwards of 80 percent of all retail transactions are being influenced, to some degree, by digital.

The retailers that are leading are those that are essentially using technology as the mortar between the bricks of their customer experience. They’re viewing technology as an essential means of removing friction from the customer experience as well as fortifying moments of experiential delight. Among the brands that I see as excelling are Amazon, Sephora, Starbucks, Wayfair – to name only a few.

MANTHAN: We hear experience becoming more important than the product – is that only for high involvement categories, while the drivers in other categories might be different?

DOUG: The consumer world is now divided into two very distinctly different and viable experiential spaces. On the one hand some brands are winning through high-fidelity experiences that are immersive, memorable and emotionally connected. Other brands are killing it in their categories with high utility experiences that are frictionless, fast, convenient and very cognitively connected – they just make sense. Both of these positions work and both are valued by consumers. The problem is, most retailers are neither high-fidelity nor high-utility and increasingly, that makes them irrelevant.

MANTHAN: Most businesses are struggling to become truly omni-channel – what’s your advice to them?

DOUG: I suppose I have two thoughts. First, if you’re still working on omni-channel, consider that Amazon, Google and others are already dealing in the realm of omni-presence, in the sense that they’ve launched technologies like Amazon’s Echo that are quietly infiltrating the majority of homes in North America and available to consumers 24/7. Secondly, I’m not a big fan of the word omni-channel. I prefer to think in terms of the customer journey with a brand and the various problems or opportunities along that journey.

If we can develop an intense and granular understanding of the consumer journey we can leverage the unique attributes of each channel to create the best possible solutions for the customer. With this insight, we can then begin to build the back and front end systems and technology architecture to bring the experience to life!

MANTHAN: Retailers that can best harness customer data will win – how far out do you see this becoming a reality?

DOUG: 200 years ago the local merchant that knew their customers most intimately won. And they gathered information about their customers by being discreet about their privacy, delivering personalized recommendations and experiences and by building trust.

Today is absolutely no different. So, yes, retailers with the best data have an advantage. But getting that data means that a clear exchange of value has to transpire. Data and privacy are no different than any other currency and consumers will spend their data with those brands and retailers that respect it and deliver clear value in exchange for it.

“I believe we’ll see a reengineering of the economic model for retail.”

MANTHAN: What’ does the ‘store of the future’ or ‘the intelligent store’ look like to you?

DOUG: 99 percent of the retail we see around us today is a relic of the 20th century. It’s retail that was built for a pre-digital era and a completely different consumer reality.

The store of the future in my opinion won’t be a “store”. It will be a space that draws the shopper into a story about their brand and their products. It will be less about the products themselves and more about productions – experiences that are interactive, immersive and fun. Technology will allow us to activate store experiences that are unique, personalized and adaptive based on unique customer preferences and needs.

Technologies built into the skeleton of the space will deliver real-time, website-like insights that will allow retailers to respond in real time to different customer groups and dynamics within the space.

Essentially, retail stores will transform from being the mini-warehouses they are today, to becoming entertainment and hospitality spaces that trade on social, physical and emotionally connected experiences. Products will come along in the jet stream across channels.

I also believe we’ll see a reengineering of the economic model for retail. Look for more retailers working directly with brands to create experiences – what I call physical media – for which they’re paid upfront, rather than being dependent on product sales.

MANTHAN: Thank you, Doug!

Build or buy a Customer Data Platform? Here’s the answer

October 29, 2018 12:07 pm No Comments October 29, 2018

A packaged, industry focused CDP provides ‘people ready data’ and quick time to value

This article is part of our CDP series. To read other articles in this series, click here.

Any marketer looking to invest in a Customer Data Platform (CDP) will have to grapple with the build vs. buy conundrum. The points of view are more entrenched in this space than any other. This is primarily because the end deliverable is a customized data platform, which resembles a services engagement. Let’s compare the build and buy options.

If you know exactly what you need, can describe it in exact requirements and have a service partner or an in-house team who can deliver to the requirements, build might be a path to consider. You will still have to wait for 6 to 12 months to get access to the solution. If you have a stop gap technology that meets your requirements while you build, or time to market is not critical consideration, you can opt for the build route.

The build option

Organizations with substantial IT budgets and large development teams with a complete range of data, design and IT skills might opt for building a CDP in-house. The people investments required to deliver a custom CDP and maintain it are considerably higher than the buy option.

Done right, there could be competitive advantages with a custom solution. However, be cognizant that at every stage there are very real challenges – scope creeps, failure to accurately define specs, cost overruns, overheads of vendor management and staff attrition.

A large-scale IT project is often lost in translation – what the customer explains is different from what the project manager understands. Further translations happen to the engineer, and then to programmer, and alas, the result ends up looking significantly different from what was envisioned.

Learn more about Manthan CDP++

The buy option

A pre-packaged CDP that is crafted for a specific industry offers the best of both worlds – quick time to value of a plug and play solution and close fit of a bespoke solution. While you would need some technical resources for the set-up and upkeep, the time and costs involved are much lower, and there are no harsh surprises.

This means you can run pilots and POCs quickly before you go all-in, you can examine and are sure that the solution really works for you. A much lower upfront cost and quick onboarding takes away risks, and still provides a solution that is tailormade for your needs.

CDP that delivers people ready data

A well-designed solution must cover scenarios and requirements that haven’t been thought of and encountered before while being cost-effective.

A CDP should serve multiple user groups that have unique needs. It should be scalable to serve data scientists who work with large data sets and require clean, well-organized data. It should be flexible to serve marketers and campaign planners who want to promptly analyze metrics and decide targeting strategies, or perform customer segmentation. Senior executives form another user group, and would typically be interested in trend reports and business dashboards.

Recognizing the distinct personas in the value chain, a differentiated CDP must have distinct interfaces and functionality for each. It’s time you addressed the unique data needs of different people.

Download CDP Handbook Now!

What goes into a Customer Data Platform Implementation

October 29, 2018 12:07 pm No Comments September 24, 2024

Milestones to track, success criteria and key stakeholders

This article is part of our CDP series. To read other articles in this series, click here.

Customer Data Platforms are in fashion, but not a trend that will fade away soon. As a pre-packaged, marketer managed system that unifies a company’s customer data to enable use cases such as personalization and contextual marketing, a CDP delivers high value to B2C businesses.

A successful CDP program requires cross-functional team effort where marketing plays a critical role in defining the program objectives and success criteria. Given the nascent stages of the technology, there is little guidance available to businesses investing in a CDP.

This implementation approach note will serve as a handy guide for B2C enterprises looking to ‘buy’ (and not build) a CDP. Read this for our view on build vs. buy.

Broadly, a CDP implementation program has three phases:

Phase 1: Business and Data Discovery

This phase involves understanding business needs, customer data and the technology landscape – along with buy-in from key stakeholders on how the unified customer record will be created and maintained in the system.

Typical activities include:

1. Planning and requirement setting

Marketing (CRM/ loyalty and marketing operations) leads this phase where the ask is to define four key program requirements

1. 1. 1. Marketing objectives – this could be improving loyalty engagement, customer retention , customer acquisition or customer engagement across touchpoints
    2. Challenges with current data – is it gaps in the data, poor quality or lack of trust in available data, or a combination of all
    3. Success criteria – metrics that business will track to validate the quality of customer data against marketing objectives
    4. Marketing campaigns – the types of campaigns that will be executed once transformed customer data is available, and how the data will be fed to marketing systems such as campaign execution tools
2. Identify data sources

This session is primarily led by data and IT engineering teams who manage customer databases. Marketing technology and analysts may also be involved if they capture customer data via surveys or campaigns. Other aspects covered in this phase require understanding the primary source for customer master, data collection, integration and verification process

1. Establish business rules for customer data unification

Based on available data and customer marketing priorities, CDP vendors work with marketing and data/ engineering to establish rules for processing and creating a comprehensive view of customer. Some of these rules could be around:

1. 1. Standardization and verification of captured customer data
  2. Prioritization of data sources for identifying customer attributes, for example, in cases where more than one value/ inconsistent data is present for a given attribute (such as more than one home store for a customer or multiple phone numbers for a customer)
  3. Establishing mechanism for de-duplicating customer records. This would require agreement on approach to run deterministic and probabilistic matches to create the golden record of a customer
2. Customer data enrichment

This consists of not only adding customer data from third party sources but also enriching customer data with analytical insights using the ‘as-is’ customer data. Examples of such enrichments are – deriving customer behavior such as a ‘discount buyer’ vs. ‘full price’ buyer – derived from historical purchases and behavior. ‘Active’ visitor or an ‘engaged’ customer based on the customers interactions on e-commerce and digital campaigns. The analytical tags maintained for every customer vary based on marketing objectives, and form critical inputs for segmenting customers to drive engagement. Marketing plays a critical role in identifying data gaps and determine how the data should be enriched.

Phase 2: Deployment

This is where the rubber meets the road. Deployment phase is vendor led (since we are talking buy option) – once critical data (such as customer master) is loaded and processed in the system, marketing is given access to the data. Other customer data activities such as enrichment of ‘nice to have’ customer attributes and cleansing of low priority customers (as defined by business) are executed in an agile manner. This ensures business can use customer data early in the cycle – they can test and validate the results and provide inputs, without having to wait till the end of implementation when it’s too late.

Data onboarding and ingestion: Getting the process and mechanisms to incorporate data sources and rules agreed upon in Phase 1 in place, to create the comprehensive customer data record
Configure data feed into downstream marketing systems: This is to ensure marketing teams have access to create customer segments, extract customer lists and data for analytical and personalization purposes
Set-up process for tracking customer metrics: This stage requires setting up a tracking and measurement process. A critical but often overlooked component of a CDP implementation is to track the health of customer data and business metrics. Information such as total customers vs. active/ inactive customers, customer retention, acquisition and conversion rates are key metrics to track to ensure the program objectives are being met
(Optional) Set-up customer engagement and campaign integrations: Not all CDPs include ‘engagement features’ – such as real-time interactions or recommendation engines that provide personalized product and offer recommendations to customers using preferred channels such as email, mobile app, online, SMS etc. B2C marketers see higher value from these optional capabilities, making it a prized component of a CDP

Phase 3: Hand over and training

The final phase in a CDP program is to ensure marketing and other customer-facing functions are trained and have access to customer data for their requirements. In addition, data/ engineering teams are trained on monitoring data ingestion and serving from the CDP on an ongoing basis.

The success of a CDP program requires establishing a governance team which comprises of marketing, IT and data engineering. Collectively, they ensure ongoing collection of new data, and verifying/ tweaking the applied rules such that they stay current and continue to meet business needs. By tracking the health of customer data and marketing metrics such as loyalty, acquisition, retention, conversion and engagement consistently, an organization can become truly customer-centric. A well-executed CDP program delivers on business goals quickly, be it controlling customer churn, increasing share of wallet or growing customer value.

Download CDP Handbook Now!

The Identity Puzzle in Customer Data Management

October 26, 2018 9:14 am No Comments October 26, 2018

This Article first appeared in MarTech

This article is part of our CDP series. To read other articles in this series, click here.

In Hindu mythology, Ravana, the great scholar and demon king, has ten heads, symbolizing his various powers and knowledge. The heads were indestructible with the ability to morph and regrow. In their battle, Rama, the warrior god, thus must go below Ravana’s heads and aim the arrow at his solitary heart to slay him for good.

In modern times, the consumer is a bit like Ravana, not in terms of his evil designs but his multiple identities. Research states that an average consumer in US today is connected to 3.64 devices. With proliferation of a host of new age devices like smart speakers, wearables, connected homes and automobiles etc., it is projected that she could be connected to as many as 20 devices in not so distant future. Like it did for Rama, this poses a clear challenge for today’s marketer – how to navigate through the maze of these devices to identify and recognize THE consumer so she can be singularly, consistently and contextually engaged across her addressable touchpoints.

Industry research suggests that only a small fraction of consumer businesses can currently accurately identify their audience – hence the advent and rapid rise of Identity management solutions that help businesses resolve the identity of their audience into individual consumer identities and profiles. The size of the Identity solutions market is estimated to grow from $900 Mn currently to over $2.6 Bn by 2022, outpacing overall marketing investments growth

A recent Winterberry research survey indicates that about 50% of consumer businesses have intensified focus and plan to increase investments on Identity solutions. While segmentation and targeting on paid media remain the predominant use cases for consumer brands, cross-device and channel personalization plus measurement and attribution are expected to become areas of focus in near future.

Identity Solutions: The past, present and future

At its core, an identity resolution solution’s job is to continuously gather audience activity data from a disparate set of data sources, platforms and services to derive a cohesive, omni-channel identity and profile of each individual audience member. However, the approach has been largely siloed so far with marketing channel specific identity platforms and strategies. CRM databases as custodians of first party customer and contact information, have been the mainstream identity platforms for direct marketing activations, primarily over email or direct mail.

With the growth of digital marketing spend, Data Management Platforms (DMPs) that store digital audience behavior data to primarily support display ad buying use cases have come into prominence. However, their relevance is now questionable with walled gardens like Facebook and Google closing doors on them. The other growing channel of influence has been mobile data platforms to support mobile device & location-based engagement.

To overcome the limitations of a disconnected, multi-channel approach that current Identity solutions like CRM databases or DMPs are constrained with, the focus is shifting to emerging modern solutions like Customer Data Platforms (CDPs) and Identity Graphs. These offer a unified, cross-touchpoint and omni-channel approach towards identity resolution and linking, enabling a fully harmonized, single view of the customer to the marketer.

The Mechanics of Identity Resolution

Identity resolution system’s key job is to continuously collect audience related data from a variety of sources and put it through an ongoing process that resolves, generates and updates this data into discrete consumer profiles, which are then used by the business for various forms of marketing or other activations.

The process comprises of 3 key steps:

Data Management – Includes ingestion of disparate set of consumer data, both identity and activity related, followed by processing and storage of this data into organized repositories.
Identity Resolution – This is a crucial and complex mix of a deterministic and probabilistic process of deriving identifiers, matching, cross-referencing and linking to unique consumer identities followed by a validation mechanism to maximize the accuracy of the resolution process.
Consumer profile generation – This associates all identifiers, attributes and activities into a harmonized, holistic Identity Graph of the consumer, an individual or a household.

What makes an Effective Identity Management Solution: 5 Mantras

Ensure the Identity system is fed with data from a wide array of data sources. Not just device activity but also the applications behind to help drill past the device, cookie or pixel and reveal the real people behind them and their behavior.
As part of data management, ensure meeting consumer privacy rights and compliance requirements of industry norms like GDPR, CCPA etc.
Identity resolution should include a consistent, rule based deterministic match process to ensure high accuracy critical to support contextual, personalized engagement in direct marketing use cases
The deterministic process must be supplemented with machine learning driven probabilistic matching to expand the data set, and meet the requirement of use cases like social media or display ad marketing that expect a wider net but relatively less 1:1 personalization
The generated consumer profile, in form of an Identity graph, while having the requisite accuracy and timeliness, should go beyond the linkages to identifiers and attributes by including the desired insights to optimally enable marketing activation use cases

Download CDP Handbook Now!

Move Over Nostradamus: Prescriptive Analytics Takes Control of Customer Engagement

October 5, 2018 12:07 pm No Comments October 5, 2018

“What next?” is a question most marketers ask themselves every day. Some base their answers on experience while others rely on the gut. However today we see a new breed of marketing professionals that are basing their decisions on evidence, thanks to new technologies and prescriptive analytics companies that help make sense of the seemingly random data.

In the increasingly competitive fashion retail industry, where little can be left to chance, while planning is critical, the course of action for every shopper must be decided in a split second to capitalize on the micro-moments.

While predictive analytics tools gives you a view of what’s coming, prescriptive analytics models tells you exactly what should be done to make the most of the situation.

The Goliath in the Room

Machine learning has taken decision-making in marketing departments to the next level. Predictions and recommendations are being made by matching customer profile with product attributes. Easy to do with a handful of products, but virtually impossible to manage when customer segments and product categories go through the roof.

In fashion, for example, a purchase goes far beyond the basic attributes of color, size, fabric, design, and price. Sticking to basic attributes such as ‘t-shirt, blue, large, polo’ isn’t going to make the sale; there are hundreds of product subtleties that need to be factored in. Placement of logo, sleeve length, the texture of buttons – these features can make or break the sale for today’s discerning customers. While these are obvious to a consumer the moment they see the item, it’s extremely tricky for the marketing analyst to get right.

Driving Customer Engagement with AI

Given the ever-multiplying number of attributes, decoding data into actionable insights is now best left to AI driven analytics. Large retailers can predict what customers are likely to buy, and then prescribe the most efficient route to close the sale.

Last year, UK based retailer ASOS announced a significant improvement in predicting Customer Lifetime Value (CLTV) in marketing through the use of AI. The retailer built a model that classifies a given customer as valuable, and potentially how valuable based on signals such as customer’s demographics, purchase history, returns history, and web and app session logs.

(Fore)Seeing Patterns in the Data

The biggest reason for using AI is its inherent ability to perceive a deeper understanding of context, customer preferences and how they make purchase decisions. AI-driven analytics enables retailers to dive deeper into consumer data, automating recommendations for customers, providing them with information that is relevant and meaningful. Gartner’s Hype Cycle for Retail Technologies, 2018 mentions cognitive expert advisors (CEAs) as a technology with high benefit, with the potential to improve customer engagement by making recommendations and aiding decisions.

Crucially, this ace up the sleeve of retailers can bring about customer loyalty by helping them find what they want quickly through a comprehensive understanding of their preferences.

Impacting Bottomline with Predictive Analytics

By accurately predicting the product-likely-to-be-bought-next, and the kind of promotion that will appeal to the given customer, retailers can immensely impact customer loyalty and drive sales.

Effective ‘next best offer’ systems rely heavily on AI and advanced analytics and can drive personalization at scale.

To learn more about how you can improve customer engagement using prescriptive analytics models, Download our guide to driving next-best-actions in Retail.

Download now

Riding the Rollercoaster: Your Approach to Customer Journey Marketing

September 18, 2018 9:02 am No Comments September 24, 2024

According to the Harvard Business Review, it can cost anywhere between five to 25 times more to acquire a new customer than to retain an existing customer. Yet brands continue to spend an average of 88% of their budget for customer lifecycle marketing in awareness strategies.

Instead of driving new customers to the top of your funnel and encouraging them to make a purchase at the end, brands that understand and address the various stages of the customer lifecycle can achieve a greater impact.

For a fashion brand, this makes a lot of sense. A customer may become aware of your brand through a pop-up sale resulting in a $40 value to the brand. But if that customer continues to be engaged, and returns five times per year and with an average spend of $40 each time, then in five years, the customer is worth $1,000.

Customer Lifecycle Marketing for Retail Success

Understanding the progression of steps a customer goes through when considering, learning, purchasing, using, and maintaining loyalty to a brand is important to your retail success.

To learn how to craft the right strategies that will get a potential customer’s attention, turn her into a paying customer and nurture her to loyalty, download our paper, “Adapting Marketing to the Customer Lifecycle in Fashion Retail”

Why Customer Data Platform Is A Must Have In Retail Today?

August 30, 2018 9:02 am No Comments August 30, 2018

With an array of options at their disposal, customers are becoming more and more demanding with each passing day. Expectations are high and if you cannot meet them, they will move on and find another brand to lavish their time and money on.

To put it simply, brands must pay more attention to its customers to deliver an experience the customer will want to return to, again and again. A thorough knowledge of the customer and his/ her preferences are therefore mandatory, as it has become essential to provide every customer with a very personalized experience.

The problem of isolated customer data

Personalization is daunting as customers today interact with brands through a variety of channels both online and offline. The collection of customer data is scattered in legacy systems – with departments such as sales, marketing and customer service holding information in silos. In addition, information contained in sources such as social media, mobile apps, email, SMS, POS, coupons, loyalty programs etc. add to the barrage of data.

Having all this data is pretty useless if the retailer is unable to connect it and utilize it to make customer-centric strategies and decisions.

Gaining a single view of your customers

Creating a unified view of the customer requires data cleansing, curation, resolution, and transformation. It is also imperative to provide seamless access of customer data to business users without requiring IT intervention.

Nothing is more important in retail today than a single view of the customer, irrespective of where she chooses to do business with your brand.

Consider a hiking enthusiast who buys shoes online and talks about how excited she is about the new shoes and her upcoming trip. She may then later visit the store to buy a jacket and nylon pants. She also redeems her loyalty points with a coalition partner brand to rent backpacking camera equipment.

The amount of information around this one customer alone is a complex amalgam of data. By factoring in more information (such as Instagram feed as to where she is hiking, her lifestyle), a clear, unified 360-degree view of the customer can help retailers and their partners know what products and content to promote to drive conversions.

Enter the Customer Data Platform: A business managed a system that brings together a company’s customer data from multiple departments and channels to enable customer data modeling and optimization of timing and nature of marketing messages.

How retail businesses benefit

Let’s look at some specifics of how a Customer Data Platform can help retailers get results:

Increased revenue from existing customers through personalized customer engagement and special offers based on individual preferences.
Improved customer satisfaction through improved customer focus – the better you know your customers, the better you’re able to anticipate and give them what they need.
Optimized ad expenditure by knowing exactly where a customer is in his purchase cycle, and intelligently targeting them with relevant messaging.

All of the above have a direct and immediate impact on both revenue and profitability of a company, making the need for customer data management real and urgent for retailers.

Data science is today a game changer in marketing, and the importance of having the right Customer Data Platform in place becomes even more apparent.

Manthan view of CDP

We define CDP as the core data infrastructure that can ingest, manage and serve data. Manthan Customer360 is a CDP that also houses analytical and data science capabilities such as segmentation, look-alike predictions, data exploration, micro-segmentation, and self-service analysis.

To learn more about how a CDP can serve your retail business beyond customer analytics – operations such as assortment planning, pricing etc., download our Customer Data Platforms handbook

Gartner’s 2018 Hype Cycle for Retail Technologies. What you need to know.

August 30, 2018 9:01 am No Comments September 30, 2024

Gartner’s Hype Cycle for Retail Technologies 2018 is out, with trends for technology leaders. This year, the Hype Cycle has identified democratized AI as a key trend – products and solutions that “blur the lines between human and machine”.

We are pleased that Manthan has found mention in 5 categories in the 2018 Hype Cycle.

AI in Retail (Rating: Transformational)
Algorithmic Retailing (Transformational)
Cognitive Expert Advisors (High)
Customer-Centric Merchandising and Marketing (High)
Algorithmic Merchandise Optimization (High)

These ratings are a testament of invention and innovation in Artificial Intelligence, Advanced Analytics, and Cloud at Manthan. And a vision, to
create context-aware, AI-powered analytics products that bring the true
power of AI to every role in your business.

Manthan’s efforts have been focussed on bringing analytics-driven
decision-making to the real user. To design sophisticated analytics
products in a manner that everyone can use them. Which we believe, is
critical to the businesses dealing with the real-time, connected customer.

We architected our products by re-interpreting 4 critical areas –
Analytics Consumption, Algorithmic Processing, Solution Engineering
and Data Management.

1. Re-interpreting Analytics Consumption

Our primary goal, as Gartner puts it, is to “blur the lines between
human and machine”. This is what led to the creation of Maya – the
world’s first AI-powered conversational agent for business analytics.

With this, analytics becomes easy to consume. In a simple,
conversational format that remembers context and processes
information in real time, based on the user’s intent and his flow of
analysis. Maya is integrated with mobile, desktop and personal assistant
devices and can be invoked anytime. It offers both general and role-based
models for analytics consumers.

Maya makes use of machine learning, deep learning, advanced analytics,
cloud computing, natural language processing (NLP) and generation
(NLG), intent analysis and context-aware computing. But all you need to
do is ask.

2. Re-interpreting Algorithmic Processing

Today’s digital business is generating millions of data points, across
sources, every day. Taking traditional hierarchical approaches to
analyze data is just not physically viable. Your solution should be able
to conduct auto-discoveries, root-cause analyses, and auto-recommend
best outcomes, based on simulations.

Manthan’s analytics platform algorithmically processes anomalies,
outliers, and exceptions to recommend actions that can achieve clear,
smart goals.

3. Re-interpreting Solution Engineering

As analytics processing becomes complex, analytics experience needs to
become intuitive. Solutions need to be designed for the real decision
makers and should embed real business contexts.

Manthan’s solutions are designed to bring together advanced analytics
and algorithmic capabilities for specific use cases across retail. Our
solutions also come with the ability to scale and to incorporate new use
cases.

4. Re-interpreting Data Management

The digital business generates much more data than the traditional one,
with new data sources emerging all the time. While some of this data
drives repeatable use cases built around standard business processes,
you also cannot lose sight of new use cases that elevate the customer
experience.

Manthan offers a full-featured data management platform that can deliver
production-grade enterprise analytics in a governed model. But at the
same time, it also drives rapid experimentation and innovation with an
architecture that can ingest and mash new data sources at scale in a data
lake architecture. This supports on-demand data-processing, giving
businesses real-time decision-making abilities.

We have what you need.

We have re-interpreted analytics delivery with AI. And an elastic
cloud infrastructure and server-less computing capabilities provide the
necessary performance and agility you need from a new age data
a platform that can deliver real-time decision-making.

Tomorrow’s technology does not require a screen in front of you. Or you
in front of a screen. It will walk with you, whispering real-time
recommendations in your ear, based on intense, granular analysis.

That’s truly democratized AI. And that’s what we have for you, today.

14 Metrics that Impact Customer Churn in Fashion Retail

August 27, 2018 8:42 am No Comments August 27, 2018

Fashion retailers have been seeing disruption to their businesses for over a decade. From digitalization and the “Amazon effect”, to subscription models and fast fashion, most fashion brands have had to operate at a cut-throat level of competition in order to just survive.

In this situation, brands are particularly impacted by the loss of a customer. Not only do they experience a direct loss of revenue, but in an industry where social influence and peer reviews have a high impact, there is a cascading effect through loss of referral business. This can, in turn, impact future customer acquisition costs and marketing efforts as well.

Knowing the metrics that impact your churn can help you better understand why your customers are disengaged. By identifying potential churners before they leave, retailers can take proactive steps to neutralize it.

Breadth of Purchase (cross-category): Breadth of purchase refers to the variety of products that a customer buys. Lower cross-category purchases can indicate customers who are not deeply engaged and may simply churn for a better price point.
Customer Complaints: Who is complaining and why? Measuring this metric can help you know how close a customer is to churning. Complaints that are common to a category or product can be early indicators of larger concerns. Additionally, mentions in poor service, in-flexible returns/ exchange or frequent complaints can indicate that a customer is looking for other options.
Feedback Scores/ NPS: Churn is closely tied to customer satisfaction, so Net Promoter Score (NPS) which measures the willingness of customers to recommend a company’s products to others can be used to understand high or low levels of customer brand loyalty.
Repeat Purchase Rate (RPR): This tells you the effectiveness of your marketing strategies and retention programs. Brands can isolate one-timers and focus their energy on making them visit again, which can yield higher gains than trying to acquire new customers. The revenue from a single repeat shopper is equal to that of around 6 new customers.
Repeat Purchase Probability: Different from the RPR, the RPP is closely related to the churn rate, as customers less likely to make another purchase are more likely to churn. In fact, studies have shown that conversion rates of repeat customers are much higher – a customer who has purchased twice in the past is about 8X more likely to convert than a first-time shopper.
Customer Lifetime Value: Customer lifetime value is the future profit your business can earn from its relationship with a customer. This is based on past purchasing behavior and their likelihood to remain engaged with your brand; it is a key metric to identify your top acquisition channels and optimize your customer interactions in a way that prioritizes your best customers.
Recency: An important and often under-rated metric, recency is the time since the last transaction. By segmenting customers on recency, you can detect the impact of marketing on purchases, and filter customers most likely to churn. This often isn’t as straight-forward as it seems, because in fashion customers don’t return at fixed intervals. However, if a customer that always visits during events such as Back-to-School or Black Friday sale doesn’t show up, it is a cause for concern.
Average order value: Average order value measures the average amount of money a customer spends per purchase or average basket value (size) per order. By understanding the basket size trend, retailers can spot anomalies in purchase behavior and identify if the customer is showing early signs of churn.
Product Reviews: Product reviews are often an indication of customer satisfaction. Consistent poor reviews could indicate that a customer is in search of other options, and also create a wider negative impact, making it harder to acquire new customers.
Profitability Per Order: In addition to business success, a high PPO can also indicate if only higher margin products and full price merchandize are being sold, giving you information on whether your tactics are successful or whether churn is eroding your profits away.
Purchase Frequency: How often the average shopper makes a purchase indicates how engaged they are. A customer returning frequently for their needs indicates you are top of mind and are better positioned to drive higher revenue from them.
Time Between Purchases (TBP): The gap between purchases within a one-year period is the time between purchases (TBP). Knowing this value can you give you insights into buying patterns and segmentation, enabling you to better understand which groups are churning.
Redemption Rate (RR): The percentage of loyalty rewards being redeemed is your redemption rate and a direct indication of your customer engagement levels. The average rewards program sees a redemption rate of 14%, and retailers that are too far off from their benchmarks need to when to act to tailor the program.
Product Returns: Like customer complaints, measuring product returns can indicate which categories are likely to make a customer churn. A high return rate from an individual can build up dissatisfaction with the brand and reduce their chances of shopping again. Additionally, social feedback on product quality shared with peer groups can have a high impact on churn.

The importance of controlling churn

Fashion retailers find it especially hard to identify customers at risk of churning – shoppers don’t buy at fixed intervals and loyalty is at an all-time low. Customers that seem to be disengaged might still be interested in your brand, just not reached their re-buy period yet.

Investing in customer retention programs can make all the difference to business growth. According to Bain & Company, increasing customer retention rates by a marginal 5% can increase profits by a whopping 95%.

It is therefore critical to study the behaviors, interactions and experiences that are driving the customer relationship, to measure churn and take effective steps to minimize it.

Manthan enables marketers to execute comprehensive churn management programs. Custom-built for the retail industry, our responsive algorithms, and AI-driven execution capabilities provide marketing teams with the insights and control they need, to proactively engage with at-risk customers.

For more information, read AI Driven Approach to Boosting Customer Retention in Retail

Author Archive

Identity Solutions: The past, present and future

The Mechanics of Identity Resolution

What makes an Effective Identity Management Solution: 5 Mantras

Customer Lifecycle Marketing for Retail Success

1. Re-interpreting Analytics Consumption

2. Re-interpreting Algorithmic Processing

3. Re-interpreting Solution Engineering

4. Re-interpreting Data Management

We have what you need.