STP for a new market: choose the customer first, then the promise

Your marketing budget for the first three months in a new city will reach a few hundred people. Who do you talk to first, and what sentence do you want them to repeat? One brand has set out its promise twice, a little over three years apart: GSM on 14 April 2023 in Vietnam (Vingroup, 2023), Green SM on 30 July 2026 in Copenhagen (Green SM, 2026). This article reads the two releases within the Green SM market-entry case, sets them against Smith (1956), and ends with a working table for anyone drafting an entry plan.

What decisions does STP contain, and what did Smith (1956) say about dividing a market?

STP contains three linked decisions: divide the market into groups, choose which group to serve, choose the position to hold in customers' minds. The library's framework poses it as a question: a small-business owner about to launch a new product on a limited budget, which customers to serve and what should they remember the brand for STP: segmentation, targeting, positioning. It lists four segmentation bases: geographic, demographic, psychographic and behavioural.

Smith is the source for the first step, and his article does not use the word "STP". He sets two strategies side by side. Product differentiation draws the varied demands of many buyers onto one offer through advertising and promotion; he calls it "the bending of demand to the will of supply" and says it may be classed as a promotional strategy (Smith, 1956, pp. 4-6). Segmentation rests on differences on the demand side and adjusts product and marketing effort more precisely to what customers require (Smith, 1956, p. 5). The differentiator seeks a layer of the market cake; the segmenter seeks one or more "wedge-shaped pieces" (Smith, 1956, p. 5).

Two points carry over to a new-market plan. The choice between the two is an overall management decision, because it shapes production as well as marketing (Smith, 1956, p. 5). And the planner's challenge is to decide, preferably in advance, the level of segmentation that can be exploited with profit (Smith, 1956, p. 6); the working table below tries to answer that with numbers.

What does Green SM say it serves and promise, in Hanoi in 2023 and Copenhagen in 2026?

Neither release names a customer group by age or income; both describe the service and the promise. The promise moves from a "new era of taxis" in 2023 to being "a trusted mobility partner" in 2026.

On 14 April 2023 in Hanoi, GSM announced it was operating Green SM Taxi and would expand to at least five provinces and cities that year. Customers could book by a hotline "just like traditional taxi services", by app, or hail cars on the road; there were two tiers, GreenCar and LuxuryCar, with a fare table. The promise was taxis that are "smart, pollutant and noise-free" and "5-star service quality" (Vingroup, 2023).

The Copenhagen release is dated 30 July 2026 and was posted on Green SM's site on 4 August 2026. Unlike many ride-hailing platforms that mainly connect riders with independent drivers, it says, Green SM directly owns and manages its fleet; fares are shown before a trip is confirmed and e-receipts are issued automatically. The CEO of Green SM Europe said that if the company earns trust, growth will follow, and the release says it is not prioritising rapid expansion (Green SM, 2026). The 2023 release, by contrast, stated a plan for five provinces and cities within the year. Both are announced intentions, not results.

From here on it is my inference, not the releases' words. Neither text has customer numbers by group, market share, a survey or post-launch results, so I draw no conclusion about performance or about how the company divided the market. In 2023, segmentation shows up in the service design: two tiers, two fare structures, with LuxuryCar about 35 per cent dearer per kilometre than GreenCar over the next 24 km (21,000 divided by 15,500 is about 1.35). In Smith's terms this resembles adjusting a product line to two sets of requirements (Smith, 1956, p. 4), but nothing shows whether the two groups were measured or assumed.

How do you move from segments to a positioning statement on a small budget?

In three steps, each with a deliverable and a cheap test. I assembled the table from the library's STP framework and Smith's article; the criteria in step two are my own.

The last test comes from Green SM: a fare shown before the trip is confirmed is checkable at once; "5-star" is not.

Worked example (invented): a city courier entering a new town

All of this example is invented: the name "Bến Gió", the town, customer counts, margins, budget and acquisition costs are made up.

Bến Gió delivers parcels within Cần Thơ and wants to enter Long Xuyên. The marketing budget for the first three months is 60 million dong, six riders are set aside, each handles 40 parcels a day, and a month is 26 days. Step 1: pick two bases, behaviour (parcels per day, selling channel) and psychology (values delivery time or low price). Three groups result:

Winning 100 per cent of each group would give a monthly margin of 280.8 million dong (A), 85.8 million (B) and 234 million (C). C is the second largest, so it looks attractive, but step 2 shows size is not enough to choose:

Payback is 60 million divided by monthly margin, assuming every customer arrives in month one and none leaves. Acquisition cost is lowest for A because I assumed a few local Facebook groups and a few trial deliveries would suffice, and highest for C because I assumed undercutting national carriers would be needed. These three figures are things to test, not things you know.

Choose A; hold B for next quarter if A works; drop C for now. A needs 5 of the 6 riders; serving both A and B in full needs 8 riders and 120 million dong, more than the team and the budget set aside. Costs set the upper limit on dividing the market, as Smith says (Smith, 1956, p. 7). Step 3, the positioning statement (invented):

For cake and food sellers in Long Xuyên who need to deliver on time to regular customers, Bến Gió is a same-day city delivery service that promises a one-hour window for every parcel. Unlike delivery services that give no time window, it refunds the delivery fee if the window is missed, because every parcel gets a text confirming the time before the rider collects it.

The "because" clause is what a customer can check on the first parcel, like the fare shown first in Copenhagen. Before trusting the numbers, run ten customers.

When should you stop dividing, and where does STP sit in an entry plan?

Stop when serving one more group pushes costs beyond what you can carry. Smith writes that the upper limit of segmentation is largely defined by production-cost considerations (Smith, 1956, p. 7); for a service, in my inference, that is the cost of keeping the promise: riders, drivers, outlets. He also notes that when extra advertising and selling in core markets yields diminishing returns, smaller fringe segments may deserve attention (Smith, 1956, p. 7); in my inference, fringes are also where a small firm can get in.

STP sits behind two larger decisions: the direction of growth (the Ansoff matrix) and the route in (market entry modes). The post on Walmart in Germany tells a new-market entry where understanding the customer went wrong.

Going deeper: what I read, and the limits

Smith (1956), Journal of Marketing 21(1), pp. 3–8: I read all six pages from a scanned copy of the full article, not from the publisher's site; check the pages against the publisher's version. The page numbers I give are the printed pages. On 3 November 2026 Unpaywall reported no open-access copy. The article is conceptual, with no data tables; his examples are cigarettes, cars and refrigerators (Smith, 1956, p. 5), that is, manufacturers. Applying it to services and small firms is my inference.

Vingroup release (14 April 2023): I read the full text through the republication on MarketScreener, labelled "unedited and unaltered", and the copy on Vietnam News; the library points to vingroup.net, but that page returned HTTP 403 when I opened it. Green SM release (30 July 2026): I read the full text on greensm.com/dk-en on 3 November 2026.

I did not read Ries and Trout (2001) or Kotler and Keller (2016), the other origins the library lists for the STP framework, so the article uses none of their wording; the four segmentation bases come from the site's framework summary. The releases are marketing statements, not evidence of segmentation research.