Open almost any ecommerce Google Ads account that has been running for a year and you’ll find the same shape: several hundred keywords, of which a dozen or so produce nearly all the revenue, a long tail that has never converted once, and a search terms report full of things nobody would want to pay for.
Search ppc keywords ecommerce and you’ll get listicles of generic terms that were never going to apply to your catalog anyway. The keywords worth funding aren’t sitting on a list somewhere. They’re in your own order history, your own site search box, and your own search terms report, and they’re specific to what you stock and who around Tampa Bay is buying it.
This is the process we use to build a keyword list for a retailer, plus the modifiers that separate buyers from browsers and the negatives that stop the budget draining into nothing.
PPC keywords for ecommerce: the four buckets
Every commercial search a shopper types falls into one of four groups, and each one deserves a different bid, a different landing page, and a different expectation.
| Bucket | What it looks like | What to expect |
|---|---|---|
| Brand (yours) | Your store name, your name plus a product | Cheapest clicks and highest conversion rate in the account. Worth defending, easy to over-credit. |
| Product and model | Exact model numbers, SKUs, part numbers, specific sizes | Low volume, high intent. Usually the most profitable non-brand traffic you’ll buy. |
| Category | "outdoor ceiling fans", "womens running shoes wide" | Volume lives here. Expensive, competitive, and where sloppy accounts lose money. |
| Problem and research | "how to stop pool algae", "best generator for a house" | Rarely converts on the first click. Better handled with content and remarketing than a direct product bid. |
The mistake almost everyone makes is running all four in one campaign on one budget. Category terms then eat the money that model-number searches would have turned into orders, because category terms get far more impressions. Split them so you can fund them separately.
Build the list from data you already own
Before you touch a keyword tool, mine four sources that describe real demand rather than estimated demand.
- Your order data. Sort last year’s orders by margin, not by units. The product that sells most is often not the product worth advertising. Start your keyword list from the top twenty by contributed profit.
- Your on-site search box. Every query typed into your own site is a person who wanted something specific enough to search for it. It’s also the fastest way to find products you should stock and terms your category pages don’t use.
- Your existing search terms report. If you have any ad history at all, this is the single best source. Google documents how to pull it in the Google Ads Help Center. Look for terms that converted but were never added as keywords, and terms with spend and zero conversions over a meaningful sample.
- Customer service email and phone notes. The way customers describe a product on the phone is the way they type it into Google. Your product page probably uses the manufacturer’s language instead.
Only then reach for a keyword tool
Use the tool to find the variants and the spelling you missed, not to invent the list. And treat the volume estimates as directional. What matters is whether a term describes something you actually sell at a price you’re willing to compete on.
Modifiers that mean somebody has a card in their hand
Intent hides in the words wrapped around the product. These are the ones worth building out deliberately.
- Model and part numbers. Nobody types a part number casually. Build these out even though the volume looks tiny.
- Size, capacity, fit, and finish. "10 x 12", "3 ton", "brushed nickel", "wide fit". A shopper who knows the spec has already made the decision.
- Buying verbs. "buy", "order", "shop", "for sale", "price".
- Availability language. "in stock", "same day", "next day", "local pickup". These are gold if you can genuinely deliver on them and poison if you can’t.
- Brand plus product. If you’re an authorized dealer for a brand, those combinations convert far better than the bare category.
- Replacement and compatibility. "replacement filter for", "fits", "compatible with". Pure repeat-purchase intent.
Modifiers that usually signal the opposite: "free", "DIY", "how to", "used", "wholesale", "job", "salary", "manual", "instructions", "repair" if you sell rather than fix, and "review" unless you have genuine review content to send them to.
Tampa changes which keywords are worth funding
If you’re a national ecommerce store, geography is mostly a bidding detail. If you also have a warehouse, showroom, or counter anywhere around Tampa Bay, it’s a competitive advantage you should be buying keywords against.
Local pickup and same-day terms
A shopper in Brandon or Carrollwood who needs something today will search for it locally rather than order and wait. Terms combining your product with "near me", "Tampa", "in stock today", or "local pickup" are cheap intent for a retailer with a physical location, and completely wasted for one without. Be honest about which you are.
Hurricane season demand
Atlantic hurricane season runs June 1 to November 30, per the NOAA National Hurricane Center. If you sell generators, fuel containers, shutters, plywood, batteries, tarps, water storage, or anything else people buy when a storm enters the forecast cone, your keyword list needs a preparedness set that sits paused most of the year and comes on fast. Two practical notes: set those campaigns up in advance rather than scrambling, and make sure your stock and shipping promises can survive the traffic, because a sold-out landing page during a spike burns budget at exactly the worst moment.
Seasonality that runs backwards from the rest of the country
Tampa’s winter-resident influx runs roughly October through April, which shifts demand for homeware, patio furniture, bikes, fishing gear, and anything else that gets bought when somebody opens up a seasonal place. Meanwhile the warm climate means pool, lawn, and outdoor categories never really get an off-season the way they do in the Northeast. If you’re using a national seasonality template for your bids, you’re fighting your own market.
How much of this you can act on depends on what you’re spending. If you haven’t set a number yet, work through our guide to how much Tampa businesses should spend on Google Ads before you build a keyword list that needs three times the budget you have.
Negative keywords are half the job
A keyword list without a negative list is a hosepipe pointed at the yard. Build these lists at account level so every campaign inherits them.
- Employment: job, jobs, career, hiring, salary, apprentice
- Free and cheap: free, freebie, cheap, discount code, coupon (unless you actually run codes)
- Research and education: what is, how to, meaning, definition, wikipedia, pdf, manual, tutorial
- Wrong condition: used, second hand, refurbished, rental, hire, repair, fix
- Wrong buyer: wholesale, bulk, distributor, supplier, dropship, if you sell to consumers
- Products you don’t carry: competitor brands you don’t stock, sizes you don’t sell, the model you discontinued
Review the search terms report weekly for the first two months and monthly after that. The list is never finished, and every hour spent on it is worth more than an hour spent rewriting ad copy.
Structure so your best keywords aren’t starved
Keyword research falls apart if the account structure lets one group swallow the budget. Keep brand in its own campaign so it can’t inflate the numbers of everything else. Put your highest-margin product and model terms in their own campaign with a budget that can’t be raided. Group category terms by margin band rather than by aisle, because your bidding decision follows the margin, not the merchandising.
On the ad side, responsive search ads take up to 15 headlines and 4 descriptions, with headlines capped at 30 characters and descriptions at 90. Use the room. Put the model number, the price band, the shipping promise, and the return policy in different headlines so the system has genuinely different things to test rather than fifteen rewordings of the same sentence.
Where Shopping fits, and where it doesn’t
For retailers, keyword-driven search ads are only part of the picture. Shopping campaigns match on your product feed rather than on keywords you choose, which means feed quality does the job that keyword research does elsewhere. In practice most retail accounts want both, with search campaigns covering the model numbers and problem-led queries that a feed handles badly. Our walkthrough of Google Shopping ads for Tampa retailers covers the feed side, and it’s worth reading alongside this one because the negative keyword list you build here applies to both.
The same goes for organic. A product term you’re paying for every month is a candidate for ecommerce SEO work, and a category page that ranks reduces what you need to bid to hold the same ground.
What retailers get wrong most often
Bidding on category terms with no margin math. If your average order is $60 with a 30 percent margin, a $4 click needs to convert far better than most category traffic does. Work out what a click can be worth before you fund it.
Sending every keyword to the homepage. Someone searching a model number should land on that product, in stock, with the price visible. Every extra click between the search and the add-to-cart button costs you orders.
Broad match with no negatives and no observation. Broad match can work, but only on top of a disciplined negative list and a search terms review you actually do.
Optimizing for cost per acquisition on a store with wildly different order values. A $30 order and a $900 order shouldn’t be treated as the same conversion. Pass revenue back into the platform and manage to return on ad spend instead.
Pausing a keyword after a week of no sales. Low-volume model-number terms need a longer window before you can say anything about them.
Ignoring the product page after the click. Keyword work stops paying if the page loads slowly, hides the shipping cost until checkout, or has no reviews. Several of the account-level errors behind this show up in our list of Google Ads mistakes Tampa businesses make.
A rough worked example
This is hypothetical, but the arithmetic is the part that matters. Say your average order value is $180 and your gross margin is 35 percent, so a completed order contributes about $63. If your site converts paid traffic at 2 percent, 100 clicks produce 2 orders and about $126 of margin. That means your ceiling is roughly $1.26 per click just to break even, before you’ve paid for anything else.
Run that sum with your own numbers before you agree to any bid. It changes the conversation from "this keyword looks relevant" to "this keyword is affordable at this conversion rate", which is the only question that matters. It also tells you immediately which categories you should be advertising and which ones you should be quietly leaving to competitors who haven’t done the math.
Where to start
- Export the last 12 months of orders and rank products by total margin contributed, not units sold. Your first ten ad groups come from the top of that list.
- Pull your on-site search queries for the last 90 days and note every term your product pages don’t currently use.
- Download your search terms report, sort by cost with zero conversions, and turn the top of that list into negatives today.
- Pick your three highest-margin products and check that each one has a landing page showing price, stock, shipping cost, and returns without a scroll.
All four cost nothing but time and will usually find money before you’ve changed a single bid. If you’d rather have someone else do the digging and the build, we do this for retail and ecommerce clients alongside ongoing Google Ads management, and you can reach the Tampa team at (813) 592-8605.
Keyword research changes shape when you sell statewide rather than locally. Our page on Google Ads management in Florida covers how we handle the wider targeting.
