# The Rival Capability-Center Read: Live, Standing Up, Exploring, or Noise

*Score one rival in one city on five public signals, assign a stage with a dated next-milestone window, and label the center's likely mandate.*

- Canonical URL: https://www.refolk.ai/guides/rival-capability-center-read
- Pillar: Market and talent intelligence
- Format: Framework
- Published: 2026-11-02
- Last reviewed: 2026-11-02
- Reading time: 23 min
- Keywords: how to tell if a competitor is opening an offshore center, competitor global capability center early signals, competitor new subsidiary registration signal, competitor hiring in new city what does it mean, detect competitor engineering hub before announcement

## Key takeaways

- Anchor every capability-center timeline to the registry registration date, never to the first job post: in India, INC-22 is due within 30 days of incorporation and INC-20A within 180 days.
- An Indian entity with no INC-20A on file after 180 days is probably dormant, and a liaison office filed through FC-1 cannot earn income, so neither counts as a center.
- In Refolk's index, India has 6,411 profiles titled site lead, site leader, center head, site head, or GCC head against 35 in Poland, so a title match alone tells you far less in India.
- In Refolk's Philippines sample, 11 of 25 site leads worked at three outsourcers, so a site lead is not evidence of a captive until the rival's own entity appears in the registry.
- A posted site-leader role points to a start roughly 3.5 to 4.5 months later, given India's 60 to 90 day notice periods and an average of about 45 days to fill a role.
- Vendor setup clocks for India put first hires at 6-10 weeks through a BOT partner's entity and 4-6 months when the rival incorporates its own, and none of them states a method.

A rival has turned up in a new offshore or nearshore city, through job posts, a new local entity, or a few profile moves, and someone has to tell leadership what it means. This guide is for strategy teams, talent-intelligence analysts, and operators sizing a market. It gives you a five-dimension score, four stages (Live, Standing Up, Exploring, Noise), a dated window for the next milestone, and a mandate label, set up so that two analysts grading the same evidence reach the same stage.

## What does this read decide, and what counts as a captive center?

This read answers three questions about one rival in one city: is it building its own center there, how far along is the build, and what work will the center do. A captive center, often called a global capability center or GCC, is an offshore or nearshore unit that the parent owns and staffs to do its own work, instead of buying that work from an outsourcer.

Three build models show up in the evidence, and you need to tell them apart:

- **Own subsidiary**: the rival incorporates a local entity and employs people on it directly.
- **BOT (build-operate-transfer)**: a partner builds and runs the team on the partner's existing entity, then transfers the team to the rival's entity later.
- **EOR (employer of record)**: a third party is the legal employer on the payslip while the person works for the rival.

The four stages this read assigns:

- **Live**: an operating entity, a named site leader, a real posting cluster, and an occupied workspace or a completed transfer.
- **Standing Up**: an entity or a partner-run pod exists, and leadership or hiring is under way.
- **Exploring**: one credible signal that hasn't yet joined up with any other.
- **Noise**: once the false positives are removed, nothing is left.

New centers are common, so "a rival appeared in Hyderabad" is not news on its own. The Zinnov-Nasscom landscape counts 2,117 GCCs across 3,728 GCC units in India, with about 110 new GCCs established there between early 2024 and late 2025. In Poland, ABSL counted 61 new centers between 1 January 2024 and 31 March 2025, creating 5,400 jobs, and 90.2% of them were driven by foreign investment.

This read stops where maturity models start. Zinnov's framework grades established centers through four stages: Outpost, Satellite, Portfolio Hub, and Transformation Hub. This guide covers the period before a center reaches that ladder, when public traces are the only evidence you have.

Know the limits before you start. Every published setup clock is India-specific and comes from vendors selling one of the build models. I cover registry mechanics for India and Poland only. For Mexico, the Philippines, and Portugal, neither the clocks nor the registry fields are established here. The scoring still works there, but you have to rebuild the dating step from the local registry.

## In what order does a captive center get built?

Most builds follow a recognisable order. The entity is incorporated, a site leader is hired, the registered-office and commencement filings land, anchor hires arrive, then come a workspace and a first delivery batch. Each step leaves a different public trace, so a rival's position in this order tells you more than any single signal does.

#### The usual build order for a captive center

1. **Entity** - Incorporation creates a CIN in India or a KRS number in Poland
2. **Site leader** - The head of site is the first senior hire
3. **Filings** - INC-22 within 30 days and INC-20A within 180 days
4. **Anchor hires** - Leadership and anchor roles, then a first pod of 5 to 15
5. **Workspace** - An operating site, which vendors place at weeks 8 to 20
6. **First batch** - A first delivery team of 15 to 20, at weeks 16 to 28

*Each stage of a captive build leaves a different public trace, so the order tells you which trace to look for next.*

The full sequence, with the clocks the sources give:

1. **Mandate and model decision.** HQ decides before anything is public. You cannot observe this step.
2. **Entity incorporation.** The legal team or a setup provider handles it. Done looks like a SPICe+ CIN (India's incorporation number) or a KRS number (Poland's court register number). Wisemonk places the follow-on statutory work, FC-GPR, EPF/ESI, and Shops Act registrations, at weeks 6 to 14.
3. **Site leader hired.** The sources agree that the head or site leader should be the first senior hire. Notice periods in India run 60 to 90 days.
4. **Registered office and commencement.** INC-22 (registered office) is due within 30 days and INC-20A (declaration of commencement of business) within 180 days. Both appear on MCA, India's company registry portal.
5. **Anchor and leadership hires.** The first four to six weeks go on leadership and anchor hires, followed by a first pod of 5 to 15.
6. **Workspace.** The facilities team sets it up. Wisemonk places it at weeks 8 to 20.
7. **First batch.** Beacon Filing places this at weeks 16 to 28, with a first delivery team of 15 to 20.

The sources disagree about the order. TMS's BOT playbook hires first and incorporates in parallel, which swaps steps 2 and 5. Beacon Filing and Codiot put the entity first. So a hiring cluster with no entity behind it doesn't rule out a center. It usually means a BOT or EOR-first build.

### Setup clocks by model

| Model | First hire | Team of 10-20 | Full operation or transfer |
|---|---|---|---|
| EOR-first | 1-2 weeks | not published | entity follows |
| BOT / managed | 6-10 weeks | 3-5 months (10-15) | transfer at 12-36 months |
| Own subsidiary | 4-6 months | 6-9 months (10-15) | 9-18 months |

These figures come from Wisemonk, Codiot, Beacon Filing, and Indigrators, and none of them states a method. The BOT transfer range is wide because the sources split: Codiot and Pipal Tree say 12 to 24 months, while Beacon Filing and Wisemonk say 18 to 36. Beacon Filing also puts a captive at 12 to 18 months from board approval to 100 employees.

One clock you can derive is the leader's start date. A 60 to 90 day notice period plus the roughly 45 days an India role takes to fill means a posted site-leader role points to a start about 3.5 to 4.5 months later. The leader's notice period, not the entity, often sets the pace of the build.

## The five evidence dimensions: what each proves and how each lies

Score each of five dimensions 0, 1, or 2: registry entity, posting cluster, leadership seat, operator or EOR footprint, and workspace. The registry gives you the date that anchors every clock. The other four show how far the build has moved and which model it is using.

| Dimension | 0 | 1 | 2 |
|---|---|---|---|
| Registry entity | none, liaison office, or dormant | incorporated, commencement pending | operating entity on file |
| Posting cluster | under 3 roles or relocations | 3+ in one category in 60 days | 10+ external roles |
| Leadership seat | none, or title at a non-rival | role posted, or leader at partner | leader on rival's entity |
| Operator / EOR | none, or one lone EOR hire | several same-function partner hires | employer switch to rival entity |
| Workspace | none, or filing address only | operating address appears | first team of 15-20 on site |

### Registry entity

**What it proves:** legal intent, the parent's identity, and a date.

**India.** Every company incorporated on or after 2 November 2018 with share capital must file INC-20A within 180 days. If it doesn't, the Registrar may treat it as not carrying on business and move to strike it off, and late fees rise to 12 times the base after 180 days. A foreign company that sets up a branch, liaison, or project office files FC-1 within 30 days, FC-4 carries the parent's details, and FC-2 reports alterations within 30 days of the change. A liaison office cannot earn income in India.

**Poland.** The KRS gives free access to the board, supervisory board, PKD activity codes, and share capital. For an sp. z o.o. (a limited liability company), it also names shareholders and their stakes. A new board member or a new address must be filed within 7 days. The full extract shows every entry since registration, including past board members, and a free keyless API publishes hourly and daily change lists.

**How it lies:** a shell entity, a liaison office, or an old entity that changed hands and looks new. Score 2 only when the entity is operating: in India, INC-20A is on file; in Poland, the KRS shows a board and activity codes that match the work.

| Event | Deadline | Where it is public |
|---|---|---|
| INC-22 registered office (IN) | 30 days after incorporation | MCA |
| INC-20A commencement (IN) | 180 days | MCA |
| FC-1 branch or liaison office (IN) | 30 days from establishment | MCA |
| KRS change (PL) | 7 days | free portal, hourly API lists |

> **Tip:** Let the Polish registry watch for you
>
> The KRS API publishes hourly and daily change lists for free, so filter them for the rival's name and parent each day instead of searching by hand. A new board member or address has to appear within 7 days of the change.

### Posting cluster

**What it proves:** how much hiring is happening, and which functions it covers. You'll need the function mix later for the mandate label.

No one has published a cluster threshold with a stated method. Two heuristics circulate. CAM calls three or more roles in the same new category within 60 days near-confirmation. Stealery uses 10 roles in a new country. I use them as the 1 and 2 boundaries because analysts already know them, not because anyone has validated them.

**How it lies:** internal relocations counted as new roles, an acquired team that arrives with postings already open, and outsourcer postings for the rival's account. Count only roles posted externally under the rival, its BOT partner, or an EOR that names the rival.

### Leadership seat

**What it proves:** someone on the ground is accountable for the site. Since the head of site is the first senior hire, a named leader on the rival's own entity is the strongest non-registry signal you can find.

**How it lies:** title matching. Generic leader titles mean very different things from country to country, as Refolk's index shows.

**183x - India's count of site-leader-titled profiles compared with Poland's**

In Refolk's index, 6,411 profiles in India carry a site-lead, site-leader, center-head, site-head, or GCC-head title, against 35 in Poland and 111 in the Philippines.

| Country | Leader-titled profiles | Leader-title share of sample | Noise or vendor share of sample |
|---|---|---|---|
| India | 6,411 | 44% Center Head | 3+ of the top 10 employers are education chains |
| Philippines | 111 | 72% Site Lead | 44% at TP, Accenture, or Wipro |
| Poland | 35 | 68% Site Lead or Leader | 2+ of the top 10 are IT vendors |

The shares come from samples of 25 profiles per country in Refolk's index. In India, "Center Head" accounted for 11 of 25, and the top employers included coaching and preschool chains such as Aakash, ALLEN, and SSVM Preschool. In the Philippines, 11 of 25 came from three outsourcers: TP with 4, Accenture with 4, and Wipro with 3. Poland's sample had 17 of 25 titled "Site Lead" or "Site Leader", at employers including Visa, Exact Sciences, WP Engine, and SpotOn, alongside the vendors Capgemini and Aspire Systems.

The practical rule is that the title only counts when the employer is right. In Poland, a new Site Lead at the rival is rare and strong evidence. In India, a title search without an employer filter mostly returns schools. [Refolk](/) lets you ask for a title, an employer, and a city in a single plain-English query, so you filter on employer first and judge the title second.

### Operator or EOR footprint

**What it proves:** which build model the rival is using. Under BOT, the rival operates on the partner's registered entity while its own incorporation runs in parallel. TMS's BOT uses EOR as the legal employment engine during the build and operate phases. Lano notes that in most countries a business can't keep hiring through an EOR once it has its own entity, so a run of profiles switching employer to the rival's new local entity dates the transfer step.

**How it lies:** one person with a sales title paid through an EOR. Shor makes the point that the employer name on the payslip doesn't make sales, management, or contract activity disappear, which is a permanent-establishment risk (the risk of creating a taxable presence in that country). Companies sometimes accept that risk for one remote hire, but it doesn't make that hire a center. Several hires in one function and one city before any entity exists is the pattern that indicates BOT or EOR-first. No one has published how to weight these signals, so I treat the footprint as evidence of the build model and of progress, not as a stand-in for the registry.

Employer switches are the hardest part of this read to find by hand, because they only show up when you compare a profile's previous employer with its current one. Refolk can return exactly that move.

I ran this search: `Center heads in Hyderabad who moved from a BOT partner to the client company's own Indian entity in the last 12 months` - [see the full result list](https://www.refolk.ai/s/5bj5hfbtr3).

*Returns profiles whose employer changed from a build-operate-transfer partner to the client's own Indian entity, which is the trace that dates a transfer.*

### Workspace

**What it proves:** the rival has committed to fixed cost in the city. Vendors place the workspace at weeks 8 to 20.

**How it lies:** a registered office proves an address for filings, not a working floor. Score 1 for an operating address, such as a KRS address change or postings that name an office rather than "remote". Score 2 only when a first delivery team, which Beacon Filing sizes at 15 to 20, shows up on profiles at that site.

## How do you assign the stage and date the next milestone?

Clear the disqualifiers first, then apply the stage gates from Live downward. The first gate the evidence clears is the stage. Every clock is anchored to the registry registration date, or to the earliest external posting if no entity exists yet.

#### Legal presence against operating footprint

Horizontal axis runs from No own entity, or borrowed to Rival's own operating entity. Vertical axis runs from Thin operating footprint to Built-out operating footprint.

| Quadrant | What it means |
| --- | --- |
| Partner-run build | Standing Up via BOT or EOR; watch for employer switches to the rival's entity |
| Live or nearly | Confirm leader seat and workspace, then date the next headcount milestone |
| Exploring or noise | Run every disqualifier; re-check postings in 60 days |
| Paper entity | Standing Up inside 180 days; past 180 with no INC-20A, treat as a shell |

*Whether the rival owns the entity and how much is operating on the ground together separate a partner-run build from a paper entity.*

| Stage | Gate (first one cleared wins) | Next-milestone window |
|---|---|---|
| Live | Registry 2, Leadership 2, Posting 2, and Workspace 1+ or Operator 2 | 100 staff at 12-18 months from board approval |
| Standing Up | Registry 1+ with Leadership or Posting 1+; or Operator 1 with Posting 1+ | INC-20A by day 180; workspace weeks 8-20; first batch weeks 16-28 |
| Exploring | Any one dimension at 1, Standing Up not met | Leader starts 3.5-4.5 months after the role is posted |
| Noise | All dimensions at 0 after disqualifiers | Re-check postings in 60 days |

Dating notes:

- **Own subsidiary, Standing Up.** Take the registration date as day zero. INC-22 is due by day 30 and INC-20A by day 180. Statutory registrations fall at weeks 6 to 14, the workspace at weeks 8 to 20, and the first batch of 15 to 20 at weeks 16 to 28.
- **BOT, Standing Up.** The next milestone is the transfer, which you'll see as profiles switching employer. Give a window of 12 to 36 months from first hires and tell leadership that the sources disagree.
- **Exploring with a posted site-leader role.** Date the leader's start at 3.5 to 4.5 months after the posting date.
- **Exploring with an entity only.** Under BOT, first hires come 6 to 10 weeks after the partner engagement. Under an own entity, they come 4 to 6 months after incorporation starts.

The vendors count their weeks from project kickoff, which comes before the registration date you can see. Anchoring to registration therefore pushes each window later than the true one. Treat the windows as late bounds: if a milestone hasn't happened by the end of its window, the build has stalled or the read is wrong.

> **Rule:** The registry date anchors the clock
>
> Never date a build from the first job post. Use the registration date in MCA or KRS, and if no entity exists, record the posting date as a provisional anchor and say so in the brief.

## How do you label the center's likely mandate?

Label the mandate by the majority function among externally posted roles, then check it against the site leader's function and the registered activity codes. Use one of three labels: cost-arbitrage operations, engineering and R&D, or regional go-to-market.

| Mandate | Posting mix | Leader function | Registry check |
|---|---|---|---|
| Engineering and R&D | software, data, research roles | engineering or R&D | software or research activity codes |
| Regional go-to-market | sales, marketing, legal roles | commercial | trading or sales activity codes |
| Cost-arbitrage operations | finance, HR, service operations | operations or shared services | business-support activity codes |

Compare against the market's base rate. In Poland, 42.6% of the 61 new centers were IT and 26.2% were R&D, so 68.8% were engineering-led. An engineering read in Poland is the default and needs little defending. A go-to-market read there is the surprise, and it needs more evidence. PageCrawl reads sales, marketing, and legal postings as a full market entry rather than just engineering talent acquisition, and that function mix changes what leadership should do with the brief. The dossier gives no base rate for India or the Philippines, so don't present one.

If no function makes up more than half the cluster, label by the leader's function and mark the mandate "mixed" in the brief. If the cluster has fewer than three roles, don't assign a mandate yet.

## How to run the read on one rival and one city

Run the eight steps in the order the build itself usually happens, so each step tells you which trace to look for next. Fill in the read card in the last section as you go.

#### The capability-center read, one rival and one city

1. **Log the trigger and date it** - Record what surfaced the rival here (job post, new entity, or profile moves), the date first seen, and the source type, kept separate from any registry date.
2. **Pull the registry record** - Search MCA or KRS for the rival and its parent, recording registration number, date, form type, and activity codes; score Registry and set the anchor date or mark "none".
3. **Check commencement and office filings** - Confirm INC-22 and INC-20A in India, flag any entity past 180 days without INC-20A, and read the KRS full extract for board, address, and history in Poland.
4. **Find the leadership seat** - Search site-leader titles in the city, keep only people employed by the rival, its BOT partner, or an EOR naming the rival, and note any open site-leader posting with its date.
5. **Count the posting cluster** - Count externally posted roles in the city over the last 60 days, group them by function, and drop relocations.
6. **Map the operator or EOR footprint** - Find profiles naming an EOR or BOT partner with the rival as client, look for switches to the rival's entity, and name the build model.
7. **Confirm the workspace** - Look for a KRS address change, a named office in postings, or a first team of 15 to 20 at one site on profiles.
8. **Assign stage, window, and mandate** - Run the seven disqualifiers, apply the gates from Live downward, compute the window from the anchor date, and label the mandate from the posting mix.

Two analysts should run steps 2 to 7 separately and only compare at step 8. If they reach different stages, the disagreement almost always comes from one dimension, and it usually means someone counted a title or a payslip that should have been disqualified.

## How the read goes wrong: false positives and dating errors

Most errors are false positives, where a title, a payslip, a vendor team, or a posting date is mistaken for a build. No public source maps each false positive to the field that rules it out. The checks below are mine, derived from the registry and profile fields described above, so treat them as a working standard and test them against your own cases.

> **Watch out:** A title is not a seat
>
> In India, a "Center Head" search mostly returns coaching centres and preschools, and in the Philippines 44% of sampled site leads sit at outsourcers. Score a leadership seat only after you have confirmed the employer.

1. **Title matching inflates leadership seats.**
   - Looks like: a "Center Head" in the city, apparently confirming a leader.
   - Rule it out: confirm the employer is the rival or a named BOT partner. In Refolk's India sample, 11 of 25 leader titles were "Center Head" and the top employers included education chains.
2. **A vendor team read as a captive.**
   - Looks like: a site lead whose profile mentions the rival.
   - Rule it out: if the employer is TP or Accenture and the rival appears only as an account, it's a vendor team. Look for the rival's own entity in the registry before scoring anything above Exploring.
3. **An EOR footprint read as an entity.**
   - Looks like: profiles listing Deel or a local EOR with the rival as client.
   - Rule it out: one person with a sales title is usually a lone remote hire, so score Operator 0. Only several hires in one function and one city count as a BOT or EOR-first build.
4. **A shell entity.**
   - Looks like: a CIN exists, or an FC-1 filing names the rival.
   - Rule it out: check for INC-20A after 180 days, and check whether the FC-1 is for a liaison office, which can't earn income. Check the NIC or PKD activity codes against the claimed work. Neither a shell nor a liaison office is a center.
5. **An acquired team.**
   - Looks like: a posting cluster appears at once, with no leader search before it.
   - Rule it out: the KRS full extract shows a board change on an existing older entity, not a new registration. Read it as an acquisition, and date it from the board change.
6. **Relocations counted as a cluster.**
   - Looks like: a sudden rise in profiles in the city at the rival.
   - Rule it out: profiles show the same people moving city with the same employer. Count only roles that were posted externally.
7. **A dating error.**
   - Looks like: a confident "the center opens in Q3" built on the first posting.
   - Rule it out: use the registration date in KRS or MCA as the anchor. Using the posting date as the incorporation date shifts every window.

Two softer failures apply to the brief more than the score. First, vendor clocks are estimates with no stated method, so present them as ranges and name where they come from. Second, the gap between first hire and announcement isn't established. One vendor says postings appear often 4 to 8 weeks before an announcement, with no method, so don't promise leadership an announcement date.

> A posting date is not an incorporation date. Anchor every clock to the registry, or say plainly that you could not.

## How to brief leadership and keep the read current

Give leadership the stage, the mandate, the anchor date, the next-milestone window, and the single piece of evidence that would move the stage. Then re-check on each source's own cadence: daily for the KRS change lists, at day 30 and day 180 for MCA filings, and every 60 days for postings.

**Rival capability-center read card**

```
Rival:
City / country:
Trigger and date first seen:
Anchor date (registry registration, or "provisional: posting date"):

Scores (0 / 1 / 2) and evidence
- Registry entity: _  | form, number, date, activity codes:
- Posting cluster: _  | external roles in last 60 days, by function:
- Leadership seat: _  | name, employer, start month, or posting date:
- Operator / EOR: _   | partner or EOR named, switches to rival entity:
- Workspace: _        | address evidence, team size at site:

Disqualifiers checked (1-7):
Build model: own subsidiary / BOT / EOR-first / unknown

STAGE: Live / Standing Up / Exploring / Noise
Next milestone and window:
Mandate: engineering and R&D / regional go-to-market / cost-arbitrage operations / mixed
What would move the stage:
Confidence limits (clocks are vendor estimates; country coverage):
```

*One card per rival per city; keep the evidence column to links and dates so a second analyst can re-grade it.*

Keep the read current on a fixed rhythm:

- **Poland:** filter the KRS API change lists for the rival and parent each day. A new board member or address must appear within 7 days.
- **India:** check MCA at day 30 for INC-22 and at day 180 for INC-20A. A missed INC-20A moves Registry to 0 and usually moves the stage down.
- **Postings:** recount every 60 days, which matches the window the cluster heuristic uses.
- **Profiles:** look for employer switches from the partner or EOR to the rival's entity every time you recount postings. The first run of switches is the transfer milestone.
- **Outside India and Poland:** before you date anything, document which fields the local registry exposes and which filing deadlines apply.

#### Before the read goes to leadership

- [ ] The anchor date comes from MCA or KRS, or is clearly marked provisional.
- [ ] Every leadership seat names an employer that is the rival, its BOT partner, or an EOR naming the rival.
- [ ] The posting count includes only externally posted roles and excludes relocations.
- [ ] Any India entity past 180 days has been checked for INC-20A.
- [ ] Any FC-1 filing has been checked for liaison-office status.
- [ ] All seven disqualifiers are recorded as checked on the card.
- [ ] The stage is the first gate cleared from Live downward.
- [ ] The next-milestone window cites its source model and is presented as a range.
- [ ] The mandate label is backed by the function split, or marked "mixed" or withheld.
- [ ] A second analyst has re-graded the evidence and reached the same stage.

When a read changes stage, keep the old card and date the change. Over a few rivals, your dated transitions become your own setup clocks for the countries the vendors don't cover. That's the only reliable way to close the gap the public sources leave.

## Frequently asked questions

### How do I tell if a competitor is opening an offshore center or just hired one remote person?

Check the employer field and the function mix. A single profile with a sales title paid through an EOR is usually a lone remote hire, not a center. Several hires in one function and one city before any entity exists suggests a BOT or EOR-first build. A captive shows up as the rival's own entity in the registry, followed by a site leader and a posting cluster on that entity.

### How early can I detect a competitor engineering hub before it is announced?

No public source has established the gap between first hire and announcement. One vendor claims postings appear 4 to 8 weeks before an announcement but gives no method behind it. In practice the registry is often your earliest dated trace: in Poland, KRS changes appear on free hourly and daily change lists, and in India the incorporation record exists before most hiring traces do.

### What does a competitor's new subsidiary registration actually tell me?

It proves legal intent and gives you a date to anchor every clock to, but it does not prove a center. In India, check that INC-20A is filed within 180 days, because an entity without it is probably dormant. Check that the filing is not a liaison office, which cannot earn income. Then look for a site leader and a posting cluster on that entity.

### How many job postings in a new city count as a real cluster?

No threshold with a stated method has been published. The heuristics in circulation are three or more roles in the same new category within 60 days, which one vendor calls near-confirmation, and ten roles in a new country, from another. I use them as score boundaries because analysts already know them, not because anyone has validated them. Count external postings only.

### Does this read work outside India and Poland?

The scoring model works anywhere, but the dated windows do not carry over automatically. Every published setup clock is India-specific and comes from setup vendors, and registry mechanics for Mexico, the Philippines, and Portugal are not established here. Before you date anything elsewhere, find out what the local registry exposes: registration date, board, activity codes, and filing deadlines.

### How do I read a rival that is building through a BOT partner?

Score it Standing Up once a pod forms under the partner, even with no rival entity yet. Under BOT the rival operates on the partner's registered entity while its own incorporation runs in parallel. The milestone to watch is profiles switching employer from the partner to the rival's local entity, which dates the transfer. Sources put transfer at anywhere from 12 to 36 months.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/guides/rival-capability-center-read*