Applying STP to OMO: Customer Needs, Touchpoints, and a Test Plan
Shusaku Yosa
STP can organize the customer and value assumptions behind OMO initiatives connecting online and offline experiences. Clear STP does not guarantee success, and complete data integration is not a universal starting requirement. Begin with one customer problem, then define the touchpoints, delivery conditions, and evidence needed to evaluate the proposed experience.
Connect STP to a touchpoint worksheet
Use segmentation to organize relevant differences, targeting to select the first problem to address, and positioning to define the promise. Only then decide what each online and store contact should provide. More connected channels are not automatically a better customer outcome.
Decision | OMO planning question | Output |
|---|---|---|
Segmentation | Which cross-touchpoint problems differ? | Observations, customer accounts, and hypotheses |
Targeting | Which problem will we address first? | Audience, selection reasons, and exclusions |
Positioning | What value will remain consistent? | Benefit and evidence of delivery |
Execution | How will contacts and owners connect? | Conditions, exceptions, and measurement |
For the general process, use a complete STP worksheet. For grouping and selection, see segmentation versus targeting. This article concentrates on applying those decisions to a connected retail experience.
Separate observed behavior from the reason behind it
Browsing a product online does not establish a wish to purchase in store. Delivery timing, stock availability, size checks, payment methods, or simple research might explain the behavior. Add customer conversations or explicit choices rather than treating a behavioral record as a complete account of motivation.
Information status | Example | How to use it |
|---|---|---|
Verified connection | An account’s order and pickup record | Track only within the confirmed scope |
Aggregate only | Total store visits | Do not assign them to individual browsing records |
Customer statement | Preferred location or collection time | Treat as a preference at that time |
Hypothesis | Browsers are likely store visitors | Test before presenting it as fact |
Not every online interaction can necessarily be connected to a store record. Document unlinked and unobserved activity so that channel totals are not mistaken for unique people or complete journeys.
Choose a need that the store can deliver against
Consider a fictional retailer planning store pickup. The initial need is to compare products online and collect a selected item from a chosen store under clear conditions. This is a worked planning example, not a real company’s performance case. Select on verified preferences and eligibility, rather than an unsupported prediction of who will visit.
Check inventory confirmation, collection location, opening hours, staff responsibilities, and storage capacity. A valuable promise that cannot be delivered may produce delays and inquiries. OpenStax’s segment criteria offer a useful way to assess reach and actionability alongside the customer benefit.
Make a promise consistent across contacts
The promise is not simply “we have an app.” In this example, it is that an eligible selected item can be collected without avoidable uncertainty under the stated conditions. Online guidance and store procedures must agree, including restrictions and what happens when something goes wrong.
Positioning guidance connects customer benefit with reasons for choice. Price remains a possible choice criterion; STP does not require every initiative to avoid price competition. If a discount encourages pickup, include its cost and the possibility that existing demand merely changed channels.
Specify contacts, exceptions, and evaluation
Stage | What to provide | Result or exception to inspect |
|---|---|---|
Comparison and request | Eligible products, store, and pickup conditions | Misunderstanding and abandoned requests |
Preparation and notification | Accurate ready-for-collection information | Preparation delays and stock failures |
Collection | Location and required identification procedure | Waiting and difficulty finding the desk |
Changes or noncollection | Clear change and storage-expiry guidance | Inquiries, cancellations, and costs |
This is a planning structure created for the fictional example. Add an owner, capacity constraint, and verified data conditions to each row. Decide whether outcomes are measured per order or per customer. Do not add channel counts and call the result a count of distinct people.
A limited pilot can reveal whether the process works before broader rollout. Record the exceptions rather than excluding every failed request from the apparent success rate. They may identify the part of the service that most needs attention.
Review data gaps, costs, and customer outcomes
Compare suitable populations and periods. Inspect completion, waiting, noncollection, inquiries, staff work, and cost. Higher sales after launch do not by themselves establish an effect: seasonality, promotions, and demand that already existed may also explain the result.
The GOV.UK measurement guidance recommends connecting purpose, hypotheses, baselines, and context. Decide what evidence would support continuation, revision, or stopping. Customer research and a limited set of verified touchpoints can support an initial test; investigate additional data connections where a specific decision requires them.
- Can the store fulfill the stated promise?
- Which people and contacts remain unmeasured or unlinked?
- Is the benefit worth the workload and expense?
- Can customers receive help when the normal path fails?
Use customer-experience value design to turn the identified difficulty into an improvement. Channel connection is a means; the customer’s task and the delivered outcome remain the basis for the next decision.