Post on 22-Jan-2018
1
MARTECH ENGINEMarketing Technology
Piotr Karwatka
MARTECH – PRACTICAL APPLICATION
2
Client acquisition• Dashboard for monitoring and
managing communication in paid media, e.g. Google AdWords, DoubleClick, Google Shopping, affiliate networks, aggregators and price comparison sites, social media;
• Centralized media plan;• Aggregation of marketing activities;• Remarketing aggregation;• Aggregation of a client acquisition
cost (actual cost);• Combining data from marketing, CRM,
call centers and other off-line sources; • Antifraud systems;• A network of dynamic landing pages; • Unified analytics - connecting tools,
e.g. Google Analytics, Gemius, CMS.
Purchasing retention• Dashboard for monitoring and
managing communication with clients in owned media, e.g. e-mail, SMS, push notification;
• Marketing automation;• Customer segmentation;• Product recommendations;• Loyalty programs;• Customer scoring (customer
assessment and valuation);• Unified analytics - connecting tools
e.g. Google Analytics, Gemius, CMS, system marketing automation.
Direct sales• Vendor dashboards for managing
communication with clients in on-line and off-line media;
• Monitoring customer health;• Cross- and up-selling web/marketing
mechanisms for use by vendors;• Predefined components for
communicating with customers, e.g. everyday brochures ready to send;
• Mechanisms of product recommendation;
• Mechanisms supporting direct sales, e.g. potential and risk customeralerts.
CRO/UX automation• Layout personalization;• Product recommendations;• Search engine personalization;• Navigation personalization;• Management dashboards for website
personalization.
MARTECH – LOGICS
3Source: Hybris
TOUCHPOINTS SITES, APPS, ADS, E-MAIL, OFF-LINE
ACQUISITION
RETENTION
CRO
DIRECT SALES
ACTIONABLE DATA
4
Purchase HistoryCRM
Cookies (behaviours on www)GA API
Social MediaSalesManago
Customer’s data:
Sales Datae-Commerce / ERP / POS
Data AggregationAlgorithms and Logic
Big Data + Reco Engine
Cloudera
Reporting
PersonalizedCommunication
Dynamic content
Marketing Automation:
Sales Dashboard
MARTECH OPEN ARCHITECTURE
5
MARTECH OPEN ARCHITECTURE
6
PERSONALIZATION AND/OR MARTECH – DEVELOPING
7
Analysis of shopping habits
Prototype of personalization
elements
Testing personalization
prototypes
Designing a dedicated MarTech solution
Implementationand integration
Goal – to detect key purchasing habits, system constraints and develop the concept of solution and project scope.
Realization – workshop, input data analysis (database analysis in the areas of trade, product and customer), IT systems analysis; preliminary technical analysis.
The effect of work –conclusions from the conducted analyses (used in marketing, sales, IT and UX) MarTech and personalization development plan, a preliminary plan of MarTechand personalization mechanisms application in the organization.
Goal – to develop the first version of personalization and Martech components (segmentation mechanisms, recommendation mechanisms, data aggregating and processing mechanisms) along with a plan of their use/ implementation.
Realization – creating concept, mockups, developing prototypes of mechanisms operating independently of the current IT system.
The effect of work – prototypes of personalization and MarTechmechanisms and a plan for testing them.
Goal – to test and optimize personalization and Martech prototypes.
Realization – research/testing,optimizing the mechanisms (conceptual work, mockups, developing prototypes of mechanisms operating independently of the current IT system).
The effect of work – tested and approved prototypes of personalization and MarTechmechanisms; revised MarTechand personalization development plan.
Goal - to design the final version of MarTech and personalization solutions, create mockups, and the implementation backlog.
Realization – creating final Axuremockups, preimplementationanalytics,
The effect of work – Axuremockups, implementation backlog, planned implementation analytics (IT and the mechanism application in the organization).
Goal - implementation of personalization and Martechmechanisms, using the gained knowledge in the current sales and marketing activities.
Realization - IT implementation carried out under the strict supervision of a MarTechengineer.
VENDOR DASHBOARD – ALERTS BY SEGMENT
8
Source: http://www.slideshare.net/RetentionGrid/your-retention-marketing-todos-for-each-customer-loyalty-segment/
Potential applications:• Detecting customers’ potential by
segmentation e.g. frequency of purchase, the time since the last purchase or purchase value;
• Preparing and/or automatic delivery of pre-defined e-mail campaigns, e.g. win-back campaigns for new customers who have not got back to a store;
• Detecting promising customer segments, working on customers using layers: an increase in purchase frequency, increasing the purchase value, reducing the time since the last purchase.
CRO – PERSONALIZED LAYOUT
9
Potential applications:• Homepage tailored to the customer's
profile (blocks, offer, navigation, pop-ups), personalization based on historical data, e.g. a logged in and not logged customer and data from external sources, e.g. Facebook;
• Dynamic website elements (blocks, pop-ups) appearing depending on the profile and behavior on the website.
VENDOR DASHBOARD – OFFER BY SEGMENT
10
Source: http://workingperson.com/, http://www.windsorcircle.com/
Potential applications:• Automatic preparation and/or sending an
e-mail message containing products and promotions tailored to customer segments or individual customers;
• Managing recommendations engine, taking into account the business logic, promotions, inventory and marketing plans.
VENDOR DASHBOARD – MARKETING AUTOMATION
11
Source: http://www.preact.com/, https://rjmetrics.com/resources/reports/ecommerce-buyer-behavior/
Potential applications:• Messages sent automatically to the
customer at a pre-planned scenario, e.g. abandoning the ordering process, abandoning a shopping cart, abandoned page (while browsing);
• A sequence of messages welcoming and introducing the client (onboarding);
• A sequence of messages reactivating or recovering the client;
• Dedicated offer of the day/week sent automatically to customers.
VENDOR DASHBOARD
12
Source: https://canopylabs.com/
Potential applications:• Specifying up-selling recommendations
(product range, time of transfer recommendations) directly at the level of individual clients;
• Specifying preferred format and frequency of contact by the sales department;
• Tracking individual user behavior (on-line, off-line);
• Detecting clients with increased risk of loosing them.
VENDOR DASHBOARD – UP–SELLING ALERT
13
Source: http://www.preact.com/, https://rjmetrics.com/resources/reports/ecommerce-buyer-behavior/
Potential applications:• Detecting customer segments with similar
shopping preferences; • Detecting clients with specific behavior,
e.g. impulsive shopping, promotion shopping, purchasing supplemental stocks of a product;
• Detecting customers interested with the selected product, product type or kind of promotion/trigger e.g. a discount coupon for free delivery.
VENDOR DASHBOARD – REORDER/REPLENISHMENT ALERT
14
Source: https://www.justrightpetfood.com/
Potential applications:• Detecting the correlation between the
next purchase and a specific product (purchase recurrence);
• Developing customer segments that are willing to renew stocks of a product;
• Automatic preparation and/or sending e-mails convincing customers to repeat the purchase.
• Managing the described communication.
ACQUISITION – MONITORING COMPETITION ACTIVITIES
15
Source: Dealavo
Potential applications:• Monitoring of prices, offers, promotional
campaigns, the scope of marketing activities by competition; daily update of data; alerts.
ACQUISITION – DATA AGGREGATION
16
Source: Hybris
Potential applications:• Aggregation and integrating data from
multiple sources, e.g. CRM, Call Center, Google Analytics, Marketing Automation system, cash system, marketing tools, etc.;
• Managing a single mediaplan and purchasing media from one panel (data integration from internal systems with marketing tools, e.g. AdServer, Marketing Automation, affiliate networks);
• Supplementing aggregated data with external data e.g. demographic or social profile, data correctness.
ACQUISITION – FEED MANAGEMENT
17
Source: Lengow
Potential applications:• Marketing management based on an offer
– emitting product ads (XML); • Promotion management in the context of
sponsored links, Google Shopping, price comparison websites, offer aggregators, affiliate networks, dynamic remarketing (product presentation), RTB (product presentation), social media (FacebookAds, Pinterest), marketplace (Allegro, eBay, Amazon and other );
• Managing pricing and promotions policy from a single panel.
CORRELATION ANALYSIS
18
Potential applications:• Detecting product + product correlation,
e.g. most frequently purchased product; • Detecting correlation between
customers/users; • Detecting correlation between behavior
on the website (visiting specific sites), and purchasing;
• Detecting correlations between stimuli/triggers and purchasing e.g. customer response to promotions;
• Detecting correlation between th time of purchase and the scale and type of purchased products;
• Detecting correlation between repeating purchase.
CORRELATION ANALYSIS
19
Potential applications:• Detecting product + product correlation,
e.g. most frequently purchased product; • Detecting correlation between
customers/users; • Detecting correlation between behavior
on the website (visiting specific sites), and purchasing;
• Detecting correlations between stimuli/triggers and purchasing e.g. customer response to promotions;
• Detecting correlation between the time of purchase and the scale and type of purchased products;
• Detecting correlation between repeating purchase.
20
EXAMPLE ANALYSIS
PERSONA ANALYSIS
21
Potential applications:• Analysis of shopping habits according to
personas defined on the basis of interviews and/or testing, e.g. promotion hunters, gift buyers, thrifty customers, novelty fans, buyers using recommendations, etc.
• Analysis of stimuli/triggers in an offer or a marketing strategy stimulating customers to action;
• Modeling triggers and a method of communication (range, scope and frequency) broken down by individual personas;
• Combining qualitative and quantitative research.
ANALYSIS OF PROBABILITY
22
Potential applications:• Construction and optimization of
probability models;• Detecting customers with the highest
likelihood of purchase recurrence; • Detecting customers most likely to be
lost; • Detecting breakthroughs in building
customer loyalty, e.g. „Starting the purchase of product X significantly increases the chance of being loyal" or "after the fifth purchase the customer becomes loyal."
SHOPPING SEQUENCE ANALYSIS
23
Potential applications:• Detecting purchase sequence in the
following aspects: product category, product brand, specific product or cart size;
• Detecting shopping preferences depending on the order of purchase.
ANALYSIS AND PREDICTION OF CUSTOMER VALUE IN TIME
24
Potential applications:• Analysis and customer segmentation
according to customer value in time detecting characteristics common to the most successful clients;
• Predicting customer lifetime value (using probability analysis);
• Detecting Pareto 20% (the best clients in terms of purchase value) and aspiring segments.
FIRST PURCHASE ANALYSIS
25
Potential applications:• Analysis of marketing activities (traffic
sources, media, campaigns, triggers/discounts, season) for generating new customers;
• Detecting marketing components responsible for bringing new customers;
• Calculating the cost of acquiring a new customer;
• Multichannel analysis (taking into account conversion attribution).
TOUCHPOINTS ANALYSIS
26
Potential applications:• Detecting key points of contact with an
offer (website, application, landing pages, marketing, off-line);
• Modifying UX/marketing so that they lead customers to the appropriate places on a website;
• Detecting and removing unwanted elements in UX/marketing.
ANALYSIS AND PREDICTION OF CUSTOMER VALUE IN TIME
27
Potential applications:• Analysis of marketing activities (traffic
sources, media, campaigns, triggers/discounts, season) for expected customer value in time, the likelihood of purchase recurrence and the likelihood of becoming a loyal customer;
• Detecting marketing components responsible for bringing the most valuable customers.
THANK YOU! QUESTIONS?
28
Piotr Karwatka (pkarwatka@divante.pl)Divante Ltd – http://divante.co