Customer & Loyalty Strategy

Customer and loyalty strategy that pays for itself

We use the purchase data you already hold to see which customers make you money, who is about to leave, and whether your membership or points programme earns its cost, then design a programme your team can measure every month.

Talk to us about your customers See the questions we answer

Illustration: member sales before and after a programme redesign

Actual sales
After the redesign
80% likely range

4.7%: Sales lift needed to break even
0.96%: True point cost as a share of sales
8%: Members likely to lapse

Illustrative data

Customer and loyalty strategy decides which customers a business should invest in, with what benefits, and how the result is measured. A membership or points programme pays for itself only when the profit on the extra sales it causes covers the cost of the points it gives on all sales.

Read the full summary

NXT Consulting Group advises retailers, restaurant groups, service businesses and consumer brands in Thailand on customer and loyalty strategy. The work covers customer segmentation from real purchase data (RFM), customer lifetime value (CLV), finding customers likely to lapse (churn), designing points and membership tiers that pay for themselves (point economics), measuring whether a promotion or bonus points lifted sales against a control group, forecasting member sales, and CRM strategy on which offer to send to whom. It starts with an audit of the member data and the true cost of the current programme, then the programme is designed or redesigned and piloted with a control group before it reaches every member. We do not sell systems or apps, so the advice is not tied to any vendor. Before NXT, our team built a CRM strategy for a restaurant group and analysed customer data in retail; under the NXT name we have run customer research for an airline and a bank.

4.7%: Sales lift needed to break even, in our 2% rebate example
RFM + CLV: Segments built from real purchase data
3–6 months: Pilot against a control group before the full launch
NDA + PDPA: Confidentiality agreement and personal-data safeguards

Programme maturity

From handing out points to a programme that earns its keep

Many businesses already run points. A programme creates more value once it separates customer groups, gives them different offers, and is measured in profit.

Points
Everyone earns the same
Spend, earn, redeem from one rewards list
Segments
Customers grouped by behaviour
Tiers built from RFM and lifetime value
Personalised
Different offers by group
Group offers; lapsing customers looked after first
Profitable
Measured in real profit
Control groups; point cost tuned every month

Questions we answer

Customer questions your data can answer

We pick the method from the business question, and say so plainly when a simpler analysis would answer it.

Which customers actually make us money?

Methods

RFM
Customer lifetime value
Cohort

You getSegments from real purchase data, with the lifetime value and profit of each

Does our points programme pay for itself?

Methods

Point economy
Break-even
Scenario

You getThe true point cost, the outstanding point liability, and the sales lift needed to break even

Which customers are about to leave?

Methods

Churn model
Survival analysis

You getA risk score for each customer, the main reasons, and who to look after first

Did the promotion or bonus points really lift sales?

Methods

Control group
Uplift
A/B test

You getThe sales the campaign actually caused, compared with customers who did not get the offer

Which offer should go to whom?

Methods

Propensity model
Next best offer

You getTargets ranked by likelihood to respond, and the offer that fits each group

What should we sell next to existing customers?

Methods

Market basket
Cross-sell

You getProducts bought together, and when to offer the next one

How we work

How a customer strategy engagement runs

01 · Diagnose
Audit the programme and data
Gather member, sales and redemption data, work out the true cost of the current programme, and say early how far the data can answer the question.
02 · Design
Design the programme
Define segments, tiers, benefits and earn rates, with a financial model showing the sales lift needed to break even.
03 · Pilot
Pilot against a control group
Run it with part of the base for 3–6 months against customers who do not get it, to measure the real lift before rolling it out.
04 · Track
Track it every month
Hand over a customer dashboard and the way to tune earn rates, and train the client team to review the numbers and adjust on its own.

Data we use

Customer data most businesses already hold

The data does not need to be perfect to start. We begin by checking its quality, and say plainly how far it can answer the question.

Purchase history from POS and ERP
Membership and CRM records
Point earning and redemption
LINE OA and app data
Campaign and promotion data
Product and reward costs
Customer satisfaction surveys
External data such as holidays and seasons

We work under a confidentiality agreement (NDA), use only the data the question needs, and recommend removing personally identifiable fields before sharing, in line with Thailand's Personal Data Protection Act (PDPA).

What you receive

What we deliver

The format depends on the question. The previews below are mock-ups of the formats, not client work.

Mock-up

Programme break-even model

Change earn rates and benefits and see the cost and the sales lift needed yourself

Mock-up

Segments and customer scores

Segment, lifetime value and lapse risk for every customer

Mock-up

Customer dashboard

Member, repeat-purchase and point-cost indicators, refreshed from your existing data

Mock-up

Member sales forecast

Forecasts with a likely range, for targets and the programme budget

Mock-up

New programme plan

Target groups, benefits, the pilot plan and indicators, on a few pages

Team experience

Customer work our team has done

Selected customer research and analytics work, under the NXT name and from the team's earlier roles

Restaurants (before NXT)

CRM strategy for Central Restaurants Group

Retail (before NXT)

Customer satisfaction benchmarked against competitors, competitive strategy, and customer data analysis to find the root cause of rising marketing costs, for CP ALL

Airline (NXT engagement)

Passenger satisfaction research and segmentation for Thai VietJet Air

Bank (NXT engagement)

Customer research and satisfaction analysis by segment for a state-owned bank

Items marked 'before NXT' are the team's experience in earlier roles at other companies, not work under the NXT name. Details are on the team profiles.

See NXT's own work

FAQ

Questions before a customer strategy engagement

What does a points programme really cost? The real cost is not the advertised rebate. It is the rebate times the cost of the rewards and the share of points actually redeemed, plus the cost of running the programme. In our article's example, a programme advertised as 2% back costs about 0.96% of sales in points, and member sales must rise about 4.7% for it to break even.

Do we need our own app first? No. Many businesses start with LINE OA or their existing POS. What matters more than an app is linking each purchase to a customer, so you can measure whether the programme really lifts sales.

How much data do we need to start? At least 12 months of purchase history linked to members is enough to segment customers and estimate their value. Less is still workable, with wider uncertainty, which we state at the start.

Does NXT sell CRM systems or apps? No. We are not tied to any vendor. If you need a new system, we help define the requirements and selection criteria so the system fits the programme you designed.

Is our customer data safe? We work under a confidentiality agreement, ask only for the data the question needs, and recommend removing names, phone numbers and other identifying fields before sharing, in line with the PDPA.

How long does it take and what does it cost? Auditing and designing a programme usually takes two weeks to three months, depending on product groups, channels and data readiness, followed by a 3–6 month controlled pilot. Scope and fee are agreed before work begins.

Insights