Every GEO portal on hashtag.org has two jobs. The first is to be findable when somebody already knows your #Name. That part is easy: people type your name, they land on you. The second job is to be findable when the visitor does not know your name yet — when they type “coffee shop,” or “real estate agent,” or “ai consultant,” or any of the thousand other phrases they would have typed into Google or Yelp ten years ago. That second job is what keywords solve, and it is the single highest-leverage decision you make as a portal owner. This article walks through exactly how keyword bidding works on hashtag.org, how to buy and renew, what to do when you get outbid, and how to think about which phrases are actually worth your money.
The mental model: an honest annual auction
A keyword on hashtag.org is a phrase you pay to rank for inside the search box. Visitors who type that phrase see portals ordered by their current bid, top-bid first. The mechanic is intentionally boring: there is one number per portal per phrase — the annual amount the portal owner is willing to pay for one year of that ranking position — and the portal with the highest number is shown first. We never inject ads. There is no separate “sponsored” lane. There is no algorithmic mystery layer that re-orders the result based on your watch history. It is an auction, and you can see the auction price.
That predictability is on purpose. Every other discovery system on the consumer internet has slowly become a black box that hides the spend and reorders the results for reasons you cannot inspect. We made hashtag.org the opposite: the dollar amount currently sitting in #1 for any phrase is shown to you, in plain English, including the exact figure it would cost (annual) to take it. If you want to be on top of “coffee shop” in your town, you will know what that costs before you click anything.
Buying a keyword (the actual flow)
Open your portal’s admin page from the Navbar grid menu, scroll to the “Keywords” section, and type the phrase you want. Two things will happen:
- We show you the current top bid for that phrase against all portals, including the incumbent owner if there is one. This is the number you have to beat to land at #1.
- We pre-fill an offer that is one cent higher than the current top bid (rounded sensibly). You can edit this; you can also add multiple phrases to the same purchase.
From here, the price you see is the annual price — not a click price, not a CPM, not a daily budget. We charge you once for the year and we tell you exactly when your renewal will run. You then choose how to pay:
- Buy with card. Standard fiat checkout through Stripe. The receipt is the receipt your accountant wants. The card is saved for the next renewal.
- Pay with Gigi credits. If you already have enough Gigi credits on file (from previous top-ups, partner payouts, or refunds), you will see a button labelled “Pay with $X.XX Gigi credits”. Click it and the keyword is yours instantly — no Stripe redirect, no second receipt. Renewals can run from credits too, with a card as the fallback.
- Pay with stablecoin (optional). If you connected a wallet, crypto checkout in USDC is available. Most portal owners do not use this; we keep it for the people who already live in Web3.
Whichever method you pick, the second the payment settles you are at #1 for that phrase. There is no “under review” queue and no editorial step. Visitors searching that phrase from that moment forward see your portal first.
Buying a single keyword vs. a bundle
You can buy multiple phrases in one transaction. Doing it that way is just convenience; the price each phrase costs is unaffected by being bundled. We recommend buying related phrases together when you can: if you are a coffee shop and you bid on “coffee”, you probably also want “espresso”, “cold brew”, and the name of your specific roastery if your customers search for it. The marginal annual cost of three more phrases at a few dollars each is usually trivial compared to losing the visitor who typed the phrase you forgot to bid on.
What “outbid” actually means
At any point, somebody else can come along, set a higher annual bid for the same phrase, and become #1. We call this being outbid. When it happens, two things change for you, and one important thing does not:
- Your portal’s position drops from #1 to #2 (or further down, if multiple competitors raise). You still appear; you are just not first.
- We email you and surface a banner inside your portal’s admin saying “outbid on <phrase> — current top bid is $X.XX/yr”. We include the exact dollar figure it would cost to retake #1 today, and a one-click “Get back to #1” button.
- What does not change: your annual subscription on that phrase. You keep paying what you bid for the rest of your year. Being outbid does not refund you; it just lowers your rank until you raise.
There is no automated bidding war. We will never raise your bid on your behalf. If you want to retake #1, you do it deliberately, with one click, at a price you can see in advance.
Get back to #1, or pay with credits
From the same admin section, the Get back to #1 button does the obvious thing: it sets your annual bid one cent above the new top, charges the difference (or the full new amount, depending on whether you are still inside your current paid year), and immediately moves you to #1. If you are paying from creditsand the balance is enough, the same UI shows a Use $X.XX credits button next to the Stripe button so you can stay in flow without another card swipe. If you are outbid on multiple phrases at once, the page lets you raise all of them with a single click; we batch the spend and tell you the total before charging.
For partial moves — you are happy with #2 on three phrases but want to retake #1 on one specific one — just hit the per-phrase button. Granular control is the default; we never bundle raises behind a single “rebid all” toggle.
The decision: should you raise or accept being outbid?
This is the only question that matters and most owners get it wrong both ways. Here is the framework we use internally when we coach new portal owners.
- What is the phrase actually worth to you? If you sell a $30 product, a phrase worth a $400/yr top bid is wildly out of proportion. If you sell a $5,000 service and one extra customer covers two years of bidding, the same $400/yr is cheap. Decide before you look at the auction price; otherwise you will anchor on the competitor.
- How much volume actually goes through that phrase? hashtag.org will eventually show search-volume hints inside the keyword admin. Until then, the rule of thumb is: phrases in your category that visitors actually type to find you (the ones your existing customers say when they describe what you do) are worth more than clever phrases your marketing person wants you to own.
- Is this a defensive bid or an offensive bid? Defensive bids protect a phrase that is core to your brand (your business name, your product line, your city + category combo). You should win those, period. Offensive bids try to grab visitors searching for a competitor or a category you are entering. Those are bargains in year one and tend to get expensive in year three; budget accordingly.
- Can you raise once and stay raised? A bid that you can comfortably renew next year is healthier than a bid that bankrupts you in 11 months. The competitor who outbid you is doing the same math; if your bid is sustainable and theirs is not, you will simply win on the next renewal even if you never raise again.
Two patterns we see in real data:
- Owners over-spend on vanity phrases (“the best coffee in $city”) and under-spend on the literal name of the thing they sell. Most search volume goes through the boring word.
- Owners take being outbid personally and rebid emotionally. The result is a two-portal arms race where both spend more than the phrase is worth. If a competitor is willing to lose money on a phrase to wound you, the right move is usually to take #2 and let them pay for the lesson.
Renewals, expiry, and what happens if you stop paying
Every keyword you buy is an annual subscription. We renew on the same date each year using your saved card or your Gigi credits balance. We will warn you 30 days, 7 days, and 24 hours before the renewal — but the renewal itself is silent if it succeeds: no email, no banner, just the same #1 you already had.
If a renewal fails, the keyword goes into a 14-day grace state. During grace, your rank does not drop and your portal still appears on the phrase, but we surface a big amber banner asking you to fix the payment method. If grace runs out without a successful charge or credit deduction, the keyword expires and is returned to the marketplace; whoever bids next wins it at the price they offer. We do not auto-renew you at a higher bid; we do not extend grace to wait for “just one more try.” You are in control, and the rules are the same for everyone.
Buying back an expired keyword is just buying it again at the new top-bid price. There is no penalty. There is no “you used to own this” surcharge. If a competitor swooped in during your expiry, you have to outbid them like everyone else.
What about Gigi credit holds?
When you start a keyword raise that runs against credits, we place an escrow hold on your balance for the upper-bound cost while the purchase is processed. If the action settles cleanly, the hold becomes a real debit; if anything fails, the hold is released back to spendable credits. You will see both the held and the spendable balances in the small Gigi credits dropdown in the nav, so the difference is never opaque.
A pricing tactic that almost always works
If you have a small budget and you cannot win the most competitive phrase outright, bid on the next-most-specific version of it. “Coffee” in a major US city is a battlefield; “coffee shop near $neighborhood” is usually not. Geographic specificity is your friend, and visitors who type a more specific phrase tend to convert at higher rates than visitors who type the generic one anyway. Layering five specific phrases at $20-50/yr each is almost always a better year-one spend than winning one generic phrase at $400/yr.
For most local businesses, the right portfolio looks like: your #Name and brand variants (defensive); the literal product or service name (the boring word); two or three geo-qualified versions (city, neighborhood, region); and one clever-but-true positioning phrase you actually use in conversation (“same-day quotes,” “the only shop with a Chemex bar,” etc.). Five phrases. Set a yearly budget you can re-up next year, and protect it.
FAQ
Can I bid against my own portal? No. You can only hold one annual bid per phrase per portal. Buying the phrase a second time just renews you.
Can I buy keywords for someone else’s portal? No. Keywords are attached to the portal that owns them. If you want to give a partner a portal, give them ownership of the portal first.
What if I bid the same as the current top bid? The earlier bid wins the tie. We display this transparently — if you tie a competitor, you sit at #2 until you raise by at least one cent.
Can I cancel mid-year? Yes — you stop the renewal. Your rank stays through the end of the paid year you already bought. We do not pro-rate refunds on keywords because it would invite gaming the auction.
Does GIGI bias toward portals that bid more? No. GIGI answers visitors using the data on the portal it is grounded in. Keyword bids only affect the search results page; once a visitor is on a specific portal, the agent’s answers come from that portal’s data, not from the auction.
Where to go next
If you have not claimed your #Name yet, do that first — the 60-second walkthrough gets you to a live portal in roughly the time it takes to read this paragraph. Once the portal is live and you have your first three or four winning phrases, it is time to make the agent actually good at the job: Train GIGI on your business in 10 minutes is the next stop. And if you are reading this because a third-party platform shadow-banned or demonetised you and you are looking for an exit, queue up Stop renting your audience after these — the playbook for moving an existing audience to a portal you actually own.