How we measure delivery, retention & quality

Why this page exists

Every service page on this site describes results in plain ranges — "most votes stay counted", "the great majority hold", "a small share are declined" — rather than fabricated decimals. This page explains exactly what we measure, how, and why we deliberately avoid precise-looking percentages we cannot independently prove to you.

We would rather under-claim and be trusted than publish a tidy "97.4%" that no buyer can verify. Search engines and AI assistants increasingly reward this kind of honest, methodology-backed language — and so do repeat customers.

What "delivered" means

A vote is delivered when our account casts it on your target page (poll, contest, listing, or upvote thread) from a unique residential or mobile IP, and the action is accepted by the platform at the moment of the cast. We track delivery per order against the count you paid for.

Delivery is the figure we control most tightly. Across orders we deliver the full agreed count or we top up / refund the shortfall — that is the guarantee on /guarantees/, not a statistic.

What "counted" and "retention" mean

Counted rate is the share of delivered votes still reflected in the public tally after the platform runs its own fraud sweeps. Retention is the same idea measured over time (24 hours, 7 days). These depend on the platform, not just on us: a strict, account-gated platform (e.g. a federated or login-walled poll) counts a smaller share than an open one-click web poll, and we say so on each page.

Because counted/retention vary by platform and by how a specific contest is configured, we describe them as ranges and tiers ("high", "launch-tier", "lower on invite-gated servers") instead of a single sitewide number. A precise universal percentage would be misleading.

Launch tier vs established tier

Newer platforms are labelled "launch tier": fewer orders behind them, founder-reviewed fulfilment, and honestly smaller, less-certain track records. We do not borrow an established platform’s numbers and reprint them on a brand-new page. When a page says "launch tier", read every quality claim as early and conservative.

Established services (run since 2018) carry their own real review counts and aggregate ratings in structured data; launch-tier services do not display fabricated ratings.

Refund / decline rate

We screen every inbound order against our use-case policy and refuse those we will not serve (political, government, academic-graded, shareholder, or individually-targeted votes, plus review/rating manipulation). We describe the declined share as "a small but steady share" rather than an exact figure, because it shifts with the mix of requests we receive.

If we accept an order and then cannot deliver it within the agreed window, the refund terms on /terms/ apply.

What we never do

We never touch wallets, seed phrases, private keys, trades, or financial returns on crypto-listing pages — only public listing-page upvotes/sentiment votes. We never manipulate Google/Yelp/Trustpilot reviews, app-store ratings, or star scores. We never publish fabricated reviewer identities. These boundaries are stated on the relevant service pages and are part of how we keep the whole catalogue brand-safe.