73criz

A Tech Reboot: Why Are Fashion Companies Building Their In House Analytics Competencies?

August 14, 2018 11:10 am No Comments August 14, 2018

Here’s why fashion businesses have moved from the traditional outsourcing model to building up their inhouse competencies:

Business Context: Unless you are the Kardashians, you understand your family better than an outsider. You know exactly what the business needs at a given time as opposed to a hired hand. When an activity is outsourced to an ‘expert’ outside the organization, they tend to make decisions without taking the business context into account. For example, a particular store might have lower sales because it had a new store manager and all the employees at that store are not aligned with his/ her objectives and practices. But an outsider looks through various forecasting tools and systems and has decided that a particular line of products, which would otherwise perform well, has to be discontinued.

Helps the business be more agile: Brand preference is no longer a thing. The choices are endless. And everything is needed here and now. But consider a business that has their decisions outsourced. Predictions for the season come in on Day 1. Orders are placed, and the production begins. When the products are displayed on Day 30, they realize that Cyan is in vogue and will be sold out by Day 35. A decision needs to be made on the spot by the store manager. But corporate has centralized operations to a team outside the country. Day 40, the store has completely sold out Cyan and is turning back customers. Day 75, Cyan is no longer in vogue, but the store is filled with Cyan pop-ups and unsold merchandise.

Quicker access to Data: The tech stack is complex, and data resides in multiple silos. The company has invested in so many tools and systems and has hired contractors from those providers as experts. To pull out a simple daily dashboard and present it in a format that is relevant to every store is a massive exercise. Even worse when every store manager wants the insights to be presented the way THEY want it. Compare this to an in-house team that has a system that unifies data from all the different systems and empowers business users to build their own dashboards, the way THEY want it, and you’ll see why outsourcing is not in fashion anymore.

Lower TCO over time: Fashion houses have long been hiring external experts, consultants, and hands to execute campaigns. But there is a problem. The problem of volumes. Executing multiple campaigns is turning out to be an expensive affair. As a shortcut, there are services offered by multiple consultancies and outsourcing companies that allow business to do ‘bare minimum’ marketing. Who bears the brunt of it? The customer. While agencies focus on each customer engagement as an activity, they tend to ignore terms like segmentation and personalization. Corporate CHQ might have contracted them to design fancy looking newsletters or trigger based SMS messages that go out in a set frequency and the customer is pounded with unrelated messages that result in churn at the end of the day. More and more fashion businesses are realizing this and are bearing the upfront cost of analytics software that helps them achieve personalization at scale. And they are now seeing the ROI with more customers engaging with the messaging and increasing spend with specific brands.

Bringing in a data-driven culture – In the world of Amazon, the fashion business is no longer just an art. A business that ignores the science part, is no longer relevant. Data-driven marketing has transformed from an innovative approach to a fundamental part of fashion marketing. Strategies are now built on insights pulled from the analysis of big data, collected through consumer interactions and engagements, to form predictions about future behaviors. This involves understanding the data you already have, the data you can get, and how to organize, analyze, and apply that data to better marketing efforts. Crafting experiences and engagements based on data has now moved to the boardroom and with access to real-time data and insights, even to the store.

For a more detailed view of what some of the leading fashion retailers are doing with their in-house analytics teams, visit our website

5 Steps to Mastering Customer Journey Focused Marketing in Retail

July 30, 2018 11:32 am No Comments July 30, 2018

Consumer-facing businesses are crippled by every-growing touchpoints and siloed systems that don’t speak to each other. A complete overhaul of these legacy technologies isn’t an option – it costs prohibitive with a long-drawn-out time to value.

Yet, they need a way to quickly connect these individual systems for an end-to-end view of the customer journey analytics. Customers today hunt down what they want at the price they want. They are open to sharing their data and being followed in their purchase journeys if they get value in return – whether it is in the form of better experience, elevated service, convenience, special offers or exclusive treatment.

For marketers, delivering on heightened customer expectations requires the use of customer journey analytics software at scale – acting on customer moments as they happen. Instead of reacting to customer-created journeys and go where they are going, you need the tools to manage, influence and even mold their routes to maximize their experience and your sales. Offering guidance, timely reminders (and perhaps a special offer) to customers who are wavering during the shopping process can yield significant business gains – reduced cart abandonment and improved conversions.

With more digital channels getting added to the mix, customer decision journeys are now maze-like.

We suggest five essential steps to set-up your customer journeys, without having to start-over:

Collect all customer data – chances are you already have this in different systems. It is important to include all browsing and transactional data, irrespective of whether the journey resulted in a purchase
Connect the existing data sources – Given the expansiveness of digital journeys and the variety and volumes of data, this isn’t likely to be easy but forms the foundation. This includes cleansing, organizing and resolving identity disputes. Given that retail business, today is as much about the seamless flow of information as it is about the flow of goods; it is important to have good data governance in place. Consider a Customer Data Platform that unifies all customer data.
Switch on the data science – Start with connecting touchpoints during a single journey in real-time; to analyze customers’ deflection points, unusual behaviors, channel preferences and the kind of information they seek during their moments with your business. Some customers see more images, some are interested in product information, others read more reviews, while some other scout for offers. These elements are important to truly understand customer behavior and underlying reasons for jumping channels and device.
Follow-up with the right messages. Now that you know where customers are in their journey and what they value, personalize interfaces such as email, website, mobile app. Customer insights, together with your expertise can help serve meaningful messages on the relevant channel that moves customers towards purchase decision or even do away with a likely hurdle as a special case (such as high shipping fee, few payment options etc.).
Visualize journeys across the entire customer lifecycle. Connect a customer’s entire relationship to know whether she is just warming up to your brand, a repeat buyer, a loyal customer or at-risk of moving to your competition. This insight can be used to nurture the relationship over time and achieve upsell and cross-sell. Machine learning and predictive models iteratively improve journey performance and assess what individual journeys she is on at present, and what impact these have on each other and on the lifecycle.

In all of this, it is critical to continuously form and test hypotheses, measure outcomes and re-calibrate tactics for maximum impact. By automatically getting data-driven answers to questions such as ‘what kind of messages will further the customer’s journey’, ‘what is the best time to communicate with a customer to incite a favourable action’, and ‘which channel is likely to be effective for a wavering customer’, marketing analysts can maximize returns and optimize spend.

Conclusion

Today’s competitive retail environment requires a Journey Analytics tool that is built for scenarios specific to retail. It can enable marketers to create impactful customer-centric campaigns that target their path-to-purchase and measure journey targeting outcomes.

It is important that customer journey analytics tools automatically identify the best paths for each behavioral segment, with technology doing the heavy lifting of picking the most-suited channel combinations (basis both customer and fashion retail context) and best moments to communicate, with the end goal of accelerating conversion. The system should automatically be able to place customers in the precise lifecycle stage and move them towards the desired stage and outcome through micro-journeys along the way.

Getting these critical components spot-on and combining them with predictive insights can help you drive revenue. The results are bound to be superior to workbench-based tools, that only allow creating pre-defined paths and cannot react in real time to customer behaviors and actions.

5 Ways Fashion Businesses Can Increase Traffic – Online And In Store

July 9, 2018 11:15 am No Comments July 9, 2018

It’s 2018. What’s the most revolutionary development in fashion over the past few years? Is it the check and plaid mania? Or maybe the shirtdress or the oversized, lightweight dress? Maybe it’s the apron and baby doll dresses that have entered the summer wardrobe (from the Halloween wardrobe)? The real answer to that question is data or much rather, the explosion of data that fashion houses are exposed to and the millions of ways that data can be interpreted and acted upon over the last few years. Shoppers these days are connected all the time, mobile ubiquity provides them the freedom to browse, research and shop where they want. You no longer need a customer to fill out a survey post their purchase, at a store, to understand who they are or what they like. You are now able to effectively build a profile of a customer using third party data sources like Instagram likes, for example, and get a pretty good understanding of their lifestyle before marketing to them.

If you looked at the fashion landscape and who’s really making the moolah, names like Amazon crops up. This is not right? They are not blue blood fashion retailer? Yet they are able to sell more fashion products than almost 40% of fashion retailers put together? They don’t have Giorgio Armani or a Coco Chanel suggesting what is the right outfit for every single customer? They don’t have a celebrity makeover team that is suggesting what color top a mother needs to wear to a soccer game? What they are good at is collecting data, deriving insights out of that data and acting on that data by putting the right product in front of the customer, at the right time, on the right channel.

You would assume that in this day and age, with all the power that data brings, increasing traffic to the store and getting customers to buy online should be fairly simple. Not really. The top challenges that fashion retailers are facing today when it comes increasing traffic are similar across the board and can be categorized into four main buckets:

They don’t understand the customer and what interests them. They are simply unable to the catch signals online and off.
Their customers don’t get context-aware, personalized communications.
They are not giving enough reasons for the customer to keep coming back. Their loyalty campaigns are limited to a quarterly newsletter blast for example.
They are unable to identify when a customer is ready to buy and lure them with offers that might attract that specific microsegment of customers.

To overcome these challenges the basic investment that’s needed for a fashion business is a Customer Data Platform (CDP) that gathers information from various sources like POS Systems, Online activity, Loyalty programs, CRM systems, third party data sources, etc. Once you have that in place, here are 5 ways you can exploit the insights and predictions to increase traffic to your store (online and off).

Customer Life-cycle Marketing – Get the basics sorted. Create segments that you can use to put a label on every one of your current/ past customers. This could be based on where the customer is in their relationship with the brand, for example: just acquired, first-purchase, repeat buyer, loyal, wavering, churn.
Micro-segmentation – Basic segments alone are not enough in the age of data. You need effective customer micro-segmentation based on lifestyle, life stage, behavioral, demographic and campaign responses to have an accurate view and understanding of each cluster. This would help you create highly relevant lists to run specific campaigns. There are some strong products out there that can help you do this using propensity modeling. You can forecast the future value of customer, rank and prioritize the customers and even allocate marketing budget using these advanced technologies. Make use of them.
Effectively recommend the next-best offer – You can do this based on affinity between categories, brands and customer segments. For example, if over 40% of 30+ women bought a silver color sunglass with white oversized shirts, that’s a good indication for the system to throw up that insight to you or automatically recommend the silver glasses as a ‘complete the look’ recommendation online. If you can get a software that can identify cross-sell opportunities, personalize content/apps based on stage automatically using AI, even better.
Path-to-purchase – NPS (Net Promoter Score) is not just a fancy word thrown around in the boardroom. It really does make sense, in every business. Identifying every customer segment’s journey/ path to purchase and chalking out all the touchpoints that led to a purchase is a key element of increasing traffic. Not only can it help you shorten the buyer’s journey but can help optimize marketing spend as well. Win-Win.
Churn management – The last hat tip is to effectively manage the customers that are about to or have churned. The lazy approach, taken by most fashion businesses, unfortunately, is to include them in a newsletter group that regularly gets updates on a sale or some branding comm. At this stage, the business has effectively given up on these customers in my opinion. The first step to effective churn management is to identify customers that are likely to churn based on visit history, purchase/ lack of, campaign response and overall engagement. You can then build rules and logic to entice them with offers and promotions that are most relevant to their micro-segment, and in today’s world, these can be done automatically by the machine!

Manthan named a Strong Performer in Forrester’s Customer Analytics Wave™ Q2 2018

June 12, 2018 6:42 am No Comments September 24, 2024

Modern CMOs are striving hard to build relationships with their customers that are deeper than those with their competitors and are transforming their customer marketing practices. They are moving away from mass marketing to targeted, hyper-relevant content, across all customer touch-points in the moment of truth. To consistently succeed with hyper-relevant personalization, CMOs are looking to establish rich capabilities for Customer Analytics, Data Management and Omnichannel Execution.

Having served retail and consumer businesses for over 14 years, Manthan brings a unique mix of deep industry experience and data science expertise that has translated into specialized customer analytics offerings for every segment from grocery to fashion, specialty and restaurants and for every format from department stores to convenience stores.

Manthan has now been recognized as a strong performer in The Forrester Wave™: Customer Analytics, Q2 2018

To assess the state of the customer analytics solutions market and see how the vendors stack up against each other, Forrester evaluated the strengths and weaknesses of the nine most significant customer analytics solution providers. These vendors were evaluated against a comprehensive set of 39 criteria, which were grouped into three high-level buckets: ‘current offering’, ‘strategy’, and ‘market presence’.

Forrester states that Customer Analytics solutions are “a new breed of analytics technology” that “perform common customer analytics techniques automatically, emancipating insights from customer data with ‘do-it-for-me’ (DIFM) capabilities”. According to Forrester, Consumer Insights professionals and their stakeholders “want easy and immediate access to insights”. Those adopting customer analytics solutions benefit from ‘Business user accessibility’, ‘speed to insights and action’ and ‘single view of the customer’. We couldn’t agree more.

A holistic understanding of the customer behavior and responding to their motivations and interests is critical for consumer facing businesses today. Manthan’s customers across retail are using Customer360 and TargetOne for use cases such as personalization, multi-channel marketing, location-based targeting, and product recommendations.

“Manthan specializes in problems unique to large retailers”, “It does a good job of addressing the three distinct personas in the analytics value chain — business users, data scientists, and data engineers.” -The Forrester Wave™: Customer Analytics Solutions, Q2 2018

According to Forrester, rapid deployment to match user needs, automatic insight identification, strong model monitoring and analytical transparency are key factors professionals should look for when considering a Customer Analytics solution.

“Manthan’s robust customer data model underlies prebuilt KPIs and predictive scores, which retailers and other direct-to-consumer brands find alluring.” -The Forrester Wave™ Customer Analytics Solutions, Q2 2018

This market is characterized by a high level of complexity with many different technologies and providers. Manthan differentiates itself by helping business users consume the insights out-of-box without having to deal with complex algorithms. Some of Manthan’s customers across the globe include Comcast, Payless, Les Schwab, Future Group, Big Y, Robinsons Supermarkets, Chedraui, Consum, Yum Brands, Charming Charlie and FairPrice. This recognition definitively puts Manthan on the Customer Analytics map, as it further elevates how it serves its customers; to understand, engage and wow their end-customers.

Advanced Analytics in The Restaurant Industry

May 17, 2018 9:40 am No Comments May 17, 2018

The fight over which restaurant gets to satisfy your appetite is surely heating up. With ~$800B sales last year, this is an exciting space to watch, especially now. Customer expectations are at an all-time high and the consumer behavior is simply changing. Before we jump in, let’s look at what makes the restaurant business so unique.

Peculiarities of the Quick Service Restaurant Industry:

The average ticket value is low: Think about it this way, if you picked up all the bills from the last 10 times you visited a pizza chain and added it up, you probably won’t go beyond $1000. Juxtapose this with an apparel retailer and you would see that their average ticket value by itself is probably more than $1000.

Visit frequency is higher and cyclical: Compared to other retailers, the restaurant industry might see the same customers 8-10 times in a period of 6 months and before the frequency drops. Guest frequency is one of the most important metrics that a restaurant analytics tracks.

Size of the meal matters: Apart from guest frequency, the moolah is made by maximizing the size of the order. A large coke with the burger? Small fries to go with that? You can try combo 1 and get an extra burger by paying $2?

Customer tastes don’t vary much: Fashion trends may change with time, but a guy who likes pizzas, likes pizzas, for a long time. (Until he signs up for personal training at the local gym at least). A BBQ enthusiast might always order the same slow-cooked pork roast sandwich as opposed to a vegan who might order a vegan sandwich, at the very same outlet. And this taste does not change much over time.

The restaurant industry is unique for sure, but if you look closer, the mandates from their C level can be boiled down to three simple goals:

Increase Meal Size
Increase Guest Frequency
Decrease Customer Lapsation

Gone are the days when oversized chicken costumes and cigarette smoking men, dressed as clowns were deployed as strategies to attain these goals. In today’s day and age, there are advanced tools, smarter analytics, and intelligible information that are helping restauranteurs devise effective strategies. With an increasingly digital-savvy customer, in a multichannel ordering environment, there is no dearth of data that is available for restaurants. The real trick is, however, to make sense of all these nuggets of information and derive insights that positively impact the Net Promoter Score (NPS).

Here are some examples of how some leading restaurant chains have put their data to work:

Identified Taste affinity clusters – A large pizza chain recently used millions of data points to arrive at 10 primary segments of customers and looked at their past purchase behavior to identify taste preferences. They identified that the large pizzas were primarily ordered by dine-in, family segment as opposed to the regular size pizzas which were typically ordered by office goers/ singles who visited the store. They even identified the times at which a certain pizza gets sold more and identified cross-sell items based on purchase behavior/ taste of a segment.

Buying behavior analysis– A burger chain took a different approach. They looked at purchase behavior across different channels to identify which menu items can be added to the combo for someone who orders through a mobile device as opposed to someone who prefers to visit the restaurant. They even used advanced analytics to get a single view of the customer by integrating their POS, mobile, web and social data to identify the customer and ensured that their messaging was consistent across all channels.

NPS and Feedback Analysis – A Chinese food chain used advanced analytics to integrate all the channels that they received feedback in (mobile, at location, social) to get a single view of the customer and layered it up with sentiment analysis. They used this data to give each customer a lapsation score which was then used to target them with unique offers depending on their lifecycle.

Store location analysis – Restaurant predictive analytics models were used by a coffee chain to identify the probability of a new store succeeding in a specific location as opposed to a location down the street. They identified pockets of demand and the model prescribed a set of potential locations in a given geographic area. They then used this data to score and rank comparable locations to determine the best location and format for the new store.

With the onset of advanced analytics in restaurant industry, the real question now is: As a customer, do they know what you are going to eat before you do?

How The Online Ecommerce Model Is Disrupting CPG

April 24, 2018 9:32 am No Comments April 24, 2018

In modern parlance, an industry-wide disruption isn’t inherently a negative thing; in fact, it’s often exactly what’s needed to raise standards, drive innovation, and move past outdated views of how certain businesses should function.

Just as it has disrupted many other elements of retail, it’s inarguably the case that the online ecommerce model (so greatly polished by juggernauts like Amazon) is rapidly changing how we approach the sale of Consumer Packaged Goods (CPG).

CPG were among the last holdouts from the old brick-and-mortar retail standard, but the times keep moving on. Let’s look at specifically how the CPG sector is changing.

Offline Retail Is Getting Connected

If you look to the top of this piece, you’ll see an Amazon Go store, described as “a new kind of store with no checkout required”. Essentially, you venture into the store, pluck the items you want from the shelves, and walk out; along the way, your Amazon account is identified, the items you take are monitored, and you are charged accordingly.

This kind of store is still very much in the prototype stage, as the potential problems abound. What if the internet connection goes down? How will theft be prevented? How will retail jobs be affected, retained, or adapted? Will people be willing to have everything they buy from a store be connected with an online account?

But there’s every reason to think it will become standard sooner or later, especially when they begin to financially incentivize it. Retail will be of the ‘order online, collect in store’ variety across the board, and consumers will have more options than ever before.

3Competition is Heating Up

The CPG sector has long been dominated by a select group of huge brands— those with the funds and distribution networks to get their goods in every major store and saturate promotional avenues to the extent that consumers would automatically trust them.

For examples, think of Coca-Cola, Heinz, or Hostess. There has always been room for store-brand and locally-sourced goods alongside them on shelves, but such goods have rarely had any realistic hope of being truly competitive with them. Now, things have changed.

These days, several factors combine to make it drastically more achievable to compete with (and in some cases outperform) top brands:

Social media makes steady growth possible on little to no marketing budget.
1. Without TV, radio or banner ads, growth used to be all but impossible. Today, word-of-mouth recommendations abound through social media, and a brand that masters the various platforms can drive sales without much of a budget.
Technology has greatly lowered the investment required to sell goods.
1. You don’t need a physical store; you can operate entirely online. You don’t need experience; you can learn as you go. Comprehensive ecommerce solutions not only make it quick and cheap to build online stores— they can also help you get products built. And it’s even possible to sell without having a store.
Meeting demand for options requires flexibility and maneuverability.
1. The larger the ship, the longer it takes to change direction. The agile nature of small companies allows them to rapidly adopt new tactics to suit shifting consumer preferences and trends.

In many ways, the playing field has been leveled, and though brand recognition is still very important, it has changed significantly from a question of “Which brands reliably meet my expectations?” to “Which brands are doing the things I like right now?”.

Subscription Goods are Flourishing

Tying into all the factors we’ve looked at thus far—and following the push towards smart and expedient retail processes—the subscription model has swiftly come to dominate a variety of CPG categories, and its popularity doesn’t seem to be waning.

Instead of heading to superstores to choose items one by one, people sign up for regular deliveries of curated products following particular themes. Healthy snacks, grooming products, varied collectibles; it feels as though there’s a subscription service for pretty much every type of consumable, comestible or degradable product out there.

It’s particularly interesting to examine because it shows just how differently we value things in the digital era. We generally know we’re not getting the best prices on items through subscription boxes, but we like the convenience, and the fun mixture of unpredictability (what exactly we’re going to get) and reliability (the rate of delivery).

Presentation Points are Moving

Go to a huge retail outlet and take a look at the attention paid to the packaging: glossy boxes, stands, leaflets and banners. Everything there has be not only designed but also produced and distributed, just like the products themselves. Many items will have complex and nigh-inaccessible plastic encasements.

Then look at the average branded delivery package. Relatively plain cardboard, perhaps with some padding inside. There will be a large logo, the company name, and possibly an adherence to the brand’s color palette, but otherwise there won’t be a great deal of difference in external appearance from one box to the next. It’s all very pragmatic.

After that, go ahead and open one of the subscription boxes; you’ll likely find some kind of added value in the form of a personalized card, a discount voucher, or a small toy. The companies know that connecting with the customer after the point of sale is important, even for relatively low-cost products. Brand relationships are very valuable in the long run.

I find it quite easy to imagine a scenario in which stores like Amazon Go become standard (as other retailers will certainly need to up their game in response) and customers browse through colorful product images on digital displays before collecting the actual items in the form of generic brown boxes holding personalised contents. It definitely provides a lot of food for thought when it comes to marketing CPG.

Ecommerce has never really been restricted to high-ticket items, but it’s only through outstanding levels of efficiency and automation that it has finally become a major disruption to the CPG world— and there’s no going back now.

For retailers to survive, and thrive, they need to understand the data behind consumer actions and optimize accordingly. Manthan’s analytics-driven solutions for everything from stock control to customer need analysis can help you adapt with the times.

The Unglamorous Truth of Fashion Retailing in 2018

April 18, 2018 6:45 am No Comments April 18, 2018

The global fashion retailing market is valued at 3 trillion dollars, 3,000 billion, and accounts for 2 percent of the world’s Gross Domestic Product (GDP). In recent years, increasing vertical integration and the relentless rise of online sales have created fundamental structural shifts in the fashion retail industry. Classical players are now realizing the potential that e-commerce brings to their business and predominantly e-com players are looking at newer ways of fulfilling customer needs. Read Prime Wardrobe. Economically, we see several trends shaping the industry, including fashion’s response to intensifying volatility, continued challenges with production hubs like China/Philippines, and the rise of urban centers. But the tectonic shift that the industry has seen in the past decade is the shift in consumer behavior. Traditionally, fashion retailers, be it high end or mass market, have relied on their products to ‘sell themselves’ and to a large extent, have relied on ATL advertising and seasonal sales to bring in the moolah. With an overcrowded market that is craving for the customer’s attention, do you think this model will sustain?

There are several players that are looking at newer, smarter ways of engaging customers. Be it virtual fitting rooms or virtual stores even, VR is definitely making its presence felt. With the forecast for 900M AR-enabled smartphones by the end of 2018, the fashion retail industry is forced to rethink how it creates, showcases and retails its products. Oh, the glamour of fashion shopping!

Current market analysis, however, suggests that almost 95% of fashion retailers are looking at solving existing challenges and are bullish on solutions that help them solve existential problems and are bearish on AR and VR. These ‘existential problems’ can be broadly classified as:

Carrying customer centric merchandise based on tighter forecasts, accurate merchandise plans, brand preferences, and shopper + location specific demand.
Fixing inventory challenges through real-time inventory visibility, effective inter-store transfers, MBQ resetting for greater sales and inventory efficiency.
Automating pricing and promotion decisions optimal price-offs that meet sales targets, maintains category profitability and predicts change in demand due to price interventions.
Effectively engaging the omnichannel customer that is influenced by trends online, prefers to read reviews, walks in store but looks for information on the mobile, wants offers, discount notifications, and communications to be personalized and makes a purchase decision anywhere and at any time.

Retail pundits have long been crying wolf about the convenience that online retailing giants are bringing to the fashion retail industry and the threat that comes with it. Well, the wolf is finally here. And it’s massive.

‘Moneyball’ing Data – A Closer Look at How Churn Models and Propensity Work

March 28, 2018 7:32 am No Comments March 28, 2018

No matter how well you understand the subject and all the formulas, equations and algorithms that go with it, having the ability to predict something that is going to happen in the future is cool!

If I had to make a pop culture reference to how someone can really control outcomes by making changes based on data, I would quote Moneyball. The movie that really made sabermetrics a household name and showed audiences how the ‘back office players’ really change the result of the game. Traditionally, a scout would assess which player has the potential to positively change the outcome of a game based on experience and ‘gut feel’. More scouts had failed at their job than the good ones that made it (with a whole lot of lucky guessing). The movie showed that if you took the human factor out of the play and looked at facts based on solid data, algorithms can make better predictions on who should be playing the next game.

As machine learning, deep learning, artificial intelligence, etc. become mainstream words that are taught in primary schools these days, it pays to fully understand how the system truly makes predictions and prescribes actions that a business should take. In this article, let’s look at how churn models and propensity to buy models can help you ‘moneyball’ your data.

First things first, to ‘moneyball’ your data, you first need to have data. It can be anything from sales data, customer demographics, visits, social profiles, customer feedback, etc. This data forms the basis for your models to get trained on is called ‘training data’. Models and algorithms are either pre-built or can be customized for a specific use case. For example, if you want to understand which segment of customers are going to churn in the next quarter, you can build a churn model which denotes the probability of a specific customer/ set of customers as a percentage. You can then get an output along the lines of: ‘Top 100 customers that going to churn in Q2 18’ and use that report to engage the customer better.

How does the model arrive at this percentage?

Much like how Darwin looked at the animals in Galapagos island to understand evolution, the model looks at the ‘training data’ to understand underlying patterns in past interactions. For example, the model could look at no. of calls made to the call center, emails sent to the helpdesk, ratings posted about products and drop in visit frequency to the site or store in the past 3 months and understand that every time a customer does all four, it is most likely that the customer is going to stop buying from you. This then gets converted into a percentage and let’s say that a specific customer did just 3 of those as opposed to all four, he gets a probability of churn % as 75%. Advanced models can also allow you to apply weight to specific attributes, like: A person who gave a low rating to a product gets a heavier weight as opposed to someone who did not visit the store in x amount of time. This weight can get you closer to predicting the likelihood of customer churn analysis. The more attributes you are adding to the churn model, the better.

So how does a propensity to buy model work? Similar to the churn model, it looks at past behavior, attributes, demographics, sales data, etc. of the best customers in your training data that you want more of. For example, there is a set of thousand customers that are your real cash cows and spend $1000+ on your merchandise every month. This becomes the protagonist that you are going to refer to and compare the rest of your training data set with. Let’s say that one of the patterns that the model detected was that majority of the customers that bought $1k+ merchandise were loyal to one specific brand in your store. This purchasing pattern becomes a base for you to start marketing to others that have bought that specific brand but are in the $700 per month bucket. (What do you market to them? Look at the basket of the $1k+). This is just one example. Propensity models can slice and dice your data to look at attributes, behavior, and patterns that might be so counterintuitive that a human can never see a connection between them.

The advent of AI and machine learning has really skyrocketed the applications of churn predictive models. Models have evolved from merely testing a hypothesis to systems that can offer prescriptions and take corrective actions. Several mounting ingredients promise to spread prediction even more pervasively: bigger data, better computers, wider familiarity, and advancing science. But the biggest evolution that churn predictive models are seeing is its democratization. Business users no longer need to depend on data scientists, analysts and IT to assist in decision making. They can now use and deploy these models with analytics providers making it super easy to drag and drop segments and ask pointed questions.

You no longer need to understand the subject and all the formulas, equations and algorithms that go with it. With the help of churn prediction model, the customer churn analysis can be done by everyone and everyone can now do the customer churn prediction to know what is going to happen in the future. And that’s pretty cool!

3 Tips to Personalize Retail Newsletters

March 21, 2018 10:43 am No Comments March 21, 2018

The power of sophisticated technology is no longer the privilege of technologists alone. It has moved into the hands of the common man. This consumerization of technology has put the customer in the driver’s seat and retailers are expected to deliver more convenience, more consistency, more collaboration and more customization than ever before. This evolution is forcing the once predictable retail industry, to make the customer as the focus of every merchandise, marketing, store and supply chain decisions.

As the industry shifts focus from pushing profitable products and goods to catering to the digitally empowered customer, customer engagement is playing a crucial role in bringing in the moolah. Today’s retailers have many channels, locations and devices to converse with their customers. Be it building a platform that supports omnichannel communication, installing indoor positioning systems that communicate in-store, empowering store associates with insights for effective clienteling, retailers are leaving no stones unturned. But what’s the low hanging fruit that every retailer HAS to get right? Email newsletters. Email newsletters are a great way to continue the conversation post a purchase, pave the way for further sales and keep your brand on top of the customer’s mind. This continued engagement has been a staple for any business and the difference between a great newsletter and one that moves into spam is, drumrolls, ‘personalization’.

Email personalization has been around for a while now and there are some rudimentary tick marks that any basic CRM can offer you. Things like: adding the first name of the customer after a hello, scheduling emails to go out at times that the user has opened emails in the past, setting up a flowchart of ‘if this then that’ messages, etc. But to drive maximum outcomes out of newsletters one must look at the customer as an individual and fully understand who s/he is as a person before marketing to them. Agreed, doing this at scale might be a challenge, but that’s what segmentation is for. The more time you spend in segmenting and micro-segmenting your target audience, the better the results of your newsletters. Let me elaborate with these three examples/ tips that you can use to personalize your newsletters.

Understand that your customer data is dynamic and evolves over time – The database to send newsletters either comes through the loyalty card/ membership data the retailer already has or acquires a ‘fresh’ database from a service provider that charges anywhere from 5c to $10 for each email id. Majority of promotions are run on these email id’s that the retailer has acquired over a period of time. If your team regularly cleanses this data, you are solving the issues of bounce rates, open rates etc. but the click rates? Click rates require you to offer up the right recommendation to the individual and the individual has to receive the most contextual offer that matches his lifestyle at that very point in time. For example, your loyalty data says that Mr. X bought $50 worth of meal kit on Jan 2016. This data has entered into your system and has been lying there for more than two years. Now your marketing team is looking to promote a newly launched, ‘fit and fab’ meal kit and Mr. X unwittingly enters into the newsletter list. But hey, he never clicked on the offer you sent him. Do we need to cleanse the DB again!!?

What really happened is that Mr. X got married to Ms. Y, in Jan 2017 and started shopping for grocery items instead of meal kits. He also increased his share of wallet to $500 per month by shopping for ration supplies for the entire week. He drops into the store every Sunday evening to pick up stuff for the rest of the week. It’s all there in your loyalty data.

Now, as a retailer who wants to personalize newsletters, would you look at individual purchase items or look at buying patterns and understand the intuitive meaning that every basket has? Micro-segmentation of your database just solves one part of the puzzle, but to really personalize offerings, you need to devise strategies and have systems in place to allow movement of contacts within different segments upon every new purchase. Your system should be able to tell you that Mr. X did not respond to the meal kit offer you sent him, remove him from the DB, look at his shopping patterns and recommend that you send him a ‘20% off on Organic tomatoes’ coupon. If you have your marketing team aligned with this thought, then its bound to increase conversions. But if you have systems that have recommendation algorithms that can do this at scale, every single time, even better!

Look at ‘like to like behavior’ of similar individuals – If you are going to feed a stray dog, are you going to give him broccoli or left-over meat? The meat of course. Now how do you know that the stray would prefer meat as opposed to broccoli? Have you observed its eating habits over a period of time? Of course not. You have a dog at home that would give you a cold stare if you fed him anything other than meat. You have learned that if my dog would not prefer broccoli, then the stray would not either. It’s a logical conclusion that you arrived at by assuming that dogs, in general, prefer meat because my dog eats meat. Extrapolate this thought to your marketing. (Ah, the things your dog teaches you!).

You have 10 items in your new product catalog that you have to market in your newsletter. Let’s take 3 different segments. Segment 1 – bought dairy and eggs. Segment 2 – bought meat, wine, and beer. Segment 3 – bought diary, eggs and chocolate sauce. Now, if one of the 10 items you were to market was a chocolate sauce, which segment would you pick? Segment 3. But based on purchase data, the baskets of segment 1 and segment 3 is almost the same. Can we try offering up a discount on the chocolate sauce for segment 1 based on the sales volume that segment 3 is seeing? Let’s label segment 3 as ‘Bakers’ and segment 3 as ‘suspected bakers’. See what happened there? You have created a microsegment based on purchase patterns and created a cohort called as ‘suspected bakers’. People can move in and out of this mailing list depending upon whether they purchase the chocolate sauce or not. Once you deploy such strategies to your database, you are effectively moving away from ‘trial and error’ marketing which cannot hold a candle to personalized marketing.

Pop quiz – which segment would you categorize as ‘weekend barbeque’?

Many messages, same database – Yes, audience exhaustion is real. The more the customer opens your emails and finds out that there is nothing interesting in there, the more your response rates and ultimately sales, suffer. It is tempting to send an exciting new product that you are launching in the store, but don’t give in! The same customer is present in your ‘bakers’, ‘suspected bakers’ and your ‘weekend barbeque’ lists. Are you going to send the same, ‘new product discount on ketchup’ three times? Surely, touch monitoring systems will detect send frequencies and alert you about potential overlaps, but they just look at time stamps against a particular id and do not recommend what message you need to send them.

This is where your team needs to do the due diligence and spend considerable time effort in creating microsegments. And stick to it. If you have identified ‘weekend barbeque’ as a cohort, you must stick to messaging that is appealing to this cohort. This messaging can be around the lifestyle that a weekend barbeque enthusiast might lead. For example, you send him recipes for killer BBQ ribs in week 1, discount on BBQ sauces on week 3, invite for a store cookout competition on week 5. If the contact does not respond to any of these, then it’s time to move him out of the cohort. But imagine the loyalty that this sort of message is bound to bring (and the sales).

You could also try segmenting based on the specific shopping behavior like ‘most likely to respond to coupons’, ‘most likely to respond to try new products’, but until you get to segmenting customers as ‘most likely to respond to a coupon on organic eggs’, you are not acing personalization.

Food and Grocery Analytics in 2018 – The New Age Realities

March 9, 2018 5:37 am No Comments March 9, 2018

Food and Grocery Retail is one of the fastest growing segments and as per an estimate, by 2022 consumers could be spending $100 Billion on online grocery alone, constituting more than 20% of the overall market, worth almost $ 1.5 trillion.

While there is a dip in disposable income of an average citizen, the daily bread and meat is not something she can cut down on. Retailers must build their strategy keeping her in focus, giving her the seamless and intuitive shopping experience, she desires. This is in line with what Gartner suggests, it is vital to reorganize merchandising and marketing around the needs of the consumer. But it is easier said than done. According to a study by FMI and Nielsen – many retailers are not prepared to meet the needs of omnichannel food shoppers, not for lack of multichannel assets or touchpoints, but for want of cohesive strategy integrating the physical stores and online business processes. They are struggling with fragmented data and are “failing to adequately share shopper data, segmentation and other consumer insights leading to missed opportunities”, the report claims. Poor forecasting ability also hinders scalability for the retailers. Most retailers find it challenging to strike a balance between demand planning and response execution and fail to reach the right customer at the right time with the right offer. Their marketing promotions are still following the traditional product-centric strategies rather than being aligned to customer choice and preference.

The new age realities are making the hitherto followed strategies and processes redundant. In today’s world, the shopper is being lured with sophisticated technologies that not only recognize them as individuals, pre-empt their shopping needs, but also help them plan their shopping and make most of the prevailing offers. How do the average Food and Grocery retailers keep pace with such sophistication? What should they do if they have to survive and thrive in the omnichannel world and make their storefront an attraction for the customer to pay repeat visits? How can they come closer to the customer and re-imagine their entire business by keeping the customer in the focus – be it merchandise, inventory, store operations or marketing? Further, how can they empower the teams on the shop floor with insights and prescriptive actions to deliver unique experiences to the customer and ensure the suppliers are always replenishing the right shelves with the right assortment?

Here is some essential and effective retail analytics solution for food industry which retailers can adopt in building the F&G store for the connected shopper by reimagining the systems for store operations, supplier collaboration, trade promotions, and marketing campaigns.

Download Whitepaper