There is a very specific moment that happens on shopping websites. You arrive with good intentions. Maybe you want headphones, a camping tent, a coffee machine, a software subscription, an insulin delivery device, a resort, a video game, or some other product that has somehow accumulated enough specifications to require its own municipal government.
You open the filters. Brand. Price. Size. Type. Rating. Compatibility. Features. Click, click, click. Beautiful. The internet has successfully reduced 847 options to 19, which is objectively better than 847. Unfortunately, you now have absolutely no idea which of those 19 you should buy.
This is where faceted search and faceted filtering do an excellent job right up until the exact moment they stop being helpful. That stopping point is one of the reasons I built Octivary.
What Is Faceted Search and Why Do Ecommerce Websites Use It?
Faceted search is the familiar filtering system used throughout ecommerce, marketplaces, directories, booking websites and other large catalogues. You have a collection of things. Those things have attributes. The customer selects attributes they care about, and the catalogue gets smaller.
For a tent, that might mean choosing a four-person capacity, three-season construction, waterproofing and a maximum price. Anything that does not meet those requirements disappears from the results. That is useful. Very useful, actually. I am not here to punt faceted filtering into a dumpster and send the dumpster rolling majestically toward the horizon. Octivary itself uses filtering concepts because traditional product filtering solves a legitimate problem.
The issue is that faceted filtering mostly answers one question: which options qualify?
That is not necessarily the question the customer is struggling with. Often the real question is which one should I choose? Those are two very different problems. One is narrowing a catalogue. The other is making a decision.
Faceted Filtering Narrows the Options. It Doesn't Necessarily Resolve the Decision.
Imagine you are shopping for a laptop. You filter for 16 GB of RAM, a certain screen size, your preferred operating system, enough storage and a reasonable price. Now there are twelve laptops remaining. Excellent, except Laptop A has better battery life, Laptop B has the nicer display, Laptop C has better ports, Laptop D weighs less, Laptop E costs less and Laptop F has much better graphics performance.
So you start opening tabs. Then more tabs. Soon Chrome has developed its own weather system.
Traditional faceted filtering treats many selected requirements relatively equally. Something either survives the filter or it does not. More sophisticated systems can obviously introduce sorting, boosts and ranges, but the familiar sidebar full of checkboxes generally struggles with something extremely important: not every preference matters equally.
Maybe battery life matters dramatically more to you than ports. Maybe you would happily accept slightly less storage for a much lighter laptop. Maybe price matters, but not enough to sacrifice the feature you actually came looking for. That is the territory Octivary's Choice Compass is intended to explore.
Before It Was a Choice Compass, It Had a Much Less Friendly Name
The original name I was using for this system was Multi-Criteria Decision Analysis, or MCDA. Technically, that made sense. Multi-Criteria Decision Analysis is about comparing options across multiple criteria while taking the importance of those criteria into account.
Unfortunately, "Multi-Criteria Decision Analysis" has all the warmth of a quarterly compliance seminar. Try casually telling somebody, "Just use the Multi-Criteria Decision Analysis interface." There is a strong chance they will suddenly remember they have somewhere else to be.
So somewhere along the way, MCDA became the Choice Compass.
The underlying system is still heavily influenced by multi-criteria decision-making, weighted scoring and ranking, but the new name describes the experience much better. You are not simply eliminating products. You are giving the system a direction.
Octivary Is Built Around Priorities, Not Just Checkboxes
A Choice Compass allows someone to tell the system what they want, but also what matters most. The sections themselves can be repositioned according to priority. Preferences inside those sections contribute to the scoring. The result is a ranking of available listings based on how closely they fit what that specific person asked for.
That changes the conversation from "Here are the products that survived your filters" to "Based on what you told us matters most, this appears to be your strongest match."
That distinction matters because reducing a catalogue is useful, but reducing doubt is considerably more interesting. The goal is not simply to make the product list smaller. The goal is to help someone understand why one option has risen above another.
The Goal Isn't Just Product Discovery. It Is Purchase Confidence.
There is a reason people sometimes spend an unreasonable amount of time researching relatively ordinary purchases. The problem is not always a lack of information. Sometimes the problem is an absolute landslide of information.
Specifications. Reviews. Reddit threads. Top-ten lists. Comparison articles. Manufacturer pages. Videos. Forums containing a user called something like TentWizard94 who has extremely strong opinions regarding aluminium poles.
I am not fully discrediting any of those things. Reviews can surface problems that specification sheets never will. Discussion threads can expose long-term quirks. Comparison articles can be extremely useful, and a genuinely good reviewer can save somebody hours.
I only think the direction of oneself should be considered just as strongly, and perhaps sometimes more strongly, than all of the influencing happening around us. We are living in an era where practically everybody is recommending something to everybody else. Best this. Top ten that. You NEED this. Five products you should NEVER buy. This changed my life. The collective internet appears to have consumed six coffees and become very concerned about which air fryer you own.
I think the world is getting a bit tired.
A Choice Compass is designed to turn some of that attention back toward the person actually making the decision. What matters to you? Which trade-offs are you willing to make? Which features actually affect your life? The goal is to reach the end of the process thinking, "Okay. I understand why this one came out on top."
Octivary Is Also a More Flexible Version of "Help Me Choose"
A number of websites already recognize that filtering alone is not enough. You will sometimes see a button labelled Help Me Choose, Find My Product, Which One Is Right for Me?, or something similar. Usually that takes the customer into a quiz or another interactive experience.
I think those tools are heading in an important direction. Octivary is, in some ways, an attempt to take that idea further.
A typical Help Me Choose quiz might ask five or ten questions in an order determined by the business. Your answers are evaluated and a product appears at the end. The customer participates, but the path has already been designed for them.
The Choice Compass instead exposes more of the decision-making process to the customer. They can prioritize criteria, change those priorities, alter selections and watch the rankings respond. Instead of saying, "Answer our questions and we will tell you what you need," I would rather say, "Tell us what matters, arrange those things according to importance, and let's see what direction your own priorities create."
The customer gets to keep their hands on the steering wheel.
You Can Even Turn Listing Text Into a Priority
One feature I added takes this idea into slightly stranger territory. A user can search for text inside listing descriptions.
Once that text tag is created, it breaks out into its own individual priority section inside the Choice Compass. That new section can then be moved anywhere in the priority order and scored just like another preference.
Suppose the structured product data does not have a dedicated field for something the customer cares about, but the product descriptions regularly mention it. Maybe the customer wants listings containing phrases such as "machine washable," "Canadian made," "works offline," "low profile," "quiet operation," or another characteristic that happens to be buried inside descriptive text.
That text search does not have to sit outside the scoring model as a disposable keyword search. It becomes part of the decision itself.
The customer can effectively say, "This thing buried inside the descriptions matters enough to become one of my priorities." I find that fascinating because product catalogues are messy little creatures. Important information does not always arrive politely organized into database columns wearing name tags.
What Happens When the Perfect Product Doesn't Exist?
This is another area where binary filtering can become awkward. Suppose somebody chooses five filters and one product matches all five. Great. But what happens if nothing matches everything?
Traditional filtering can produce the ecommerce equivalent of someone closing the service window: 0 results. The customer carefully described what they wanted and the website responded by releasing a tumbleweed.
Preference-based ranking allows for a different approach. Perhaps no product satisfies every criterion, but Product A satisfies the customer's three highest priorities and misses two things they ranked lower. Product B satisfies four preferences but fails the one criterion sitting right at the top.
Those products should not necessarily be treated equally. Real purchasing decisions contain compromise. Good decision tools should be able to deal with compromise rather than exploding because the mythical perfect toaster does not exist.
There Is Also an Unseen Half of the Choice Compass
The visible Choice Compass is the part customers interact with: rearranging priorities, selecting preferences, searching text and watching listings move. But there is another side to this idea that businesses may find considerably more interesting.
Choice Compasses can optionally record decision events.
A listing gets promoted. Another gets demoted. A customer moves battery life above price. A feature is selected. A preference is removed. A text tag is created. The rankings shift again. Those movements can become structured events rather than disappearing into the digital void the moment the user closes the page.
That opens an entirely different kind of product research.
Most businesses can tell what somebody eventually purchased. Analytics can tell you what pages they visited, where they clicked and where they dropped off. A Choice Compass can potentially reveal something different: why the rankings changed while the customer was deciding.
If hundreds of customers repeatedly promote products after moving "easy to clean" toward the top, that might tell a manufacturer something. If customers frequently demote a particular product when portability becomes important, that might tell them something too. If a feature the business assumed was crucial repeatedly sinks toward the bottom of people's priorities, well, the data has entered the meeting.
That kind of information could eventually help companies create better products, reorganize product lines, improve merchandising and get the right products in front of the right people.
An Interactive Choice Tool That Doesn't Need to Burn Through AI Tokens
There is also something slightly unfashionable about Octivary that I actually like. The core Choice Compass does not require an AI model to sit behind every interaction burning tokens every time somebody drags a priority upward.
It is an interactive decision system based on structured product data, scoring and user input.
AI can absolutely be useful around systems like this, especially for data normalization, categorization, enrichment and other supporting tasks. But the actual act of ranking a catalogue according to a person's preferences does not inherently need a language model generating a brand-new answer every time someone checks a box.
That means businesses can potentially provide a highly interactive recommendation experience without tying every customer interaction directly to AI token usage.
When optional decision-event tracking is enabled, those same interactions can also create a new form of structured product research data. Instead of only asking, "What did people buy?", a business can begin exploring a much more interesting question: "What did people value while deciding what to buy?"
That information could later help manufacturers build better products, help retailers improve their product mix, and help businesses match the right people with the right products more effectively.
Collaboration Compasses Make the Decision Even More Interesting
Octivary is also experimenting with collaborative Choice Compasses, and that functionality is featured live on the site.
This is where two people's priorities can participate in the same decision.
Imagine two people choosing a resort together. One cares deeply about restaurants, quiet rooms and beaches. The other wants water sports, nightlife and enough activities to construct an itinerary requiring project-management software.
The goal is not simply to let the second person's preferences become weaker because they joined second. Both people's opinions should retain equal weight while the system evaluates the combined direction.
That opens possibilities for couples, families, teams, purchasing committees and any other situation where a decision has more than one human attached to it. Which, unfortunately for spreadsheets everywhere, happens quite often.
This Works Best for Option-Centric Products
I have started thinking about certain products and services as option-centric.
An option-centric product is something where the buying decision depends on multiple meaningful characteristics, and different customers can reasonably prioritize those characteristics differently.
A Choice Compass is probably unnecessary for a shop selling three nearly identical wooden spoons. We do not need MCDA to determine the destiny of the spoon.
But consider camping tents. Capacity matters. Weight matters. Weather resistance matters. Setup style matters. Vestibules matter. Packed size matters. Seasonality matters. Materials matter. Price matters. Those priorities then change dramatically depending on whether somebody is hiking twelve kilometres into the wilderness or reversing an SUV directly beside the campsite and unloading enough equipment to establish a minor principality.
That is an option-centric product.
The same pattern appears all over the place. A business using an Octivary-style Choice Compass could include:
- Ecommerce product catalogues
- SaaS plans and software products
- Travel and hospitality businesses
- Tourism and experience directories
- Equipment manufacturers
- Electronics retailers
- Outdoor recreation products
- Health and accessibility products
- Beauty and personal-care products
- Home improvement products
- Sporting equipment
- Education and course catalogues
- Subscription services
- B2B product catalogues
- Marketplaces and directories
The deciding factor is less about the industry and more about the decision. If customers routinely arrive asking, "Which one is right for me?", then there may be an interesting problem to solve.
There Is One Thing a Choice Compass Cannot Guarantee
Once somebody establishes a direction using a Choice Compass, I still encourage them to check around. There is something Octivary cannot reliably guarantee from structured product attributes alone: durability and longevity.
A product can look spectacular on paper and still develop the structural integrity of wet cereal after eighteen months.
That is where reviews, long-term testing, discussion threads and owner experiences become extremely valuable. Use the Choice Compass to establish direction, then investigate the finalists for the things that are difficult to quantify in a catalogue.
Does it last? Do parts break? Does the software become abandoned? Does the hinge eventually stage a revolt? Does the manufacturer actually support it?
The system should help you find the product that best fits what you value. It should not pretend it owns a time machine.
When You Are Searching for Someone to Build Better Ecommerce Filters
This part is specifically for the business owner, ecommerce manager, manufacturer or extremely caffeinated product manager who arrived here after searching for things like advanced ecommerce product filtering, interactive product filter for website, faceted search development, custom product finder, guided selling software, product recommendation tool, ecommerce filtering solution, faceted navigation, preference-based product recommendations, interactive product selector, or perhaps the magnificently desperate "how do I help customers choose between too many products?"
Hello. That last one is basically my favourite problem.
If you are already looking for a developer or technology company to build faceted search for your website, there is one question worth asking before automatically commissioning another sidebar full of checkboxes: do my customers simply need help narrowing the catalogue, or do they need help making the final decision?
Sometimes faceted filtering is absolutely enough. Sometimes the customer just needs size medium, colour blue and a maximum price of $40. Done. Everybody can go home.
But if your products contain dozens of meaningful attributes, if shoppers constantly compare similar options, if your sales team repeatedly answers "Which one should I get?", or if your catalogue contains trade-offs that depend heavily on the individual customer, then filtering may only be solving the first half of the problem.
Adding another eighteen checkboxes does not magically turn a filter into a decision-support system. At some point we have simply constructed a very organized wall of checkboxes.
What I Want Octivary to Become
Octivary started with a fairly simple obsession: choices are getting complicated.
The internet solved the problem of access remarkably well. We can find almost anything. Now we have the opposite problem. We can find everything.
Everything has variants. Every variant has features. Every feature has a premium tier. Every premium tier has a comparison chart. Somewhere around tab thirteen, the joy of having options quietly packs its belongings and moves to another province.
I want Octivary to live in the space between discovery and certainty.
Traditional search helps you find options. Faceted search helps you reduce options. Reviews and comparison content help you investigate options. A Choice Compass attempts to help you understand which option fits your own priorities best.
Behind that customer experience is another opportunity: structured information about how people actually navigate difficult decisions, which features promote or demote products, how priorities shift, where compromise happens and what different customers genuinely value.
For an option-centric business, that could become much more than an ecommerce filter. It could become a product finder, a guided-selling interface, a recommendation system, a customer research tool and a new way of understanding the relationship between products and the humans attempting to choose between them.
Because sometimes the customer doesn't need another filter.
They need a compass.