01The headline finding
Out of every 100 Colombian SMBs that appear on Google Maps, 64 have no website of their own, and of the 36 that do, 12 have a site that is dead, broken, or returning 404. When a potential customer looks them up on Google, that customer lands on an Instagram with three posts, a Facebook page with no phone number, or nothing at all.
That means 76.7% of the Colombian SMBs in our sample have no working web presence. The gap is not in advanced things (mobile, SEO, schema, performance). The gap is in the basics: existing at all.
Headline finding
Of 5,305 Colombian SMBs on Google Maps, deduplicated by name and city: 3,434 have no site (64.7%) and 633 have a registered site that is dead or unreachable (an additional 11.9%). Net no-working-presence total: 4,067 businesses, or 76.7%.
02The national picture
The baseline numbers, before slicing by city or sector:
No URL of their own on file. They exist on Google Maps with an address, a phone number, and sometimes Instagram, but no website.
The site responds 200 OK and shows real content. This is the real number that supports purchase decisions.
A domain is on file, but the site returns 404, a server error, a timeout, or a blank page.
It helps to define "dead site". It is not an ugly site. It is a site that a real customer typing the URL sees as broken: an expired domain, a downed server, an unpaid WordPress hosting bill, a page that says This site can't be reached. For the customer, that is worse than no site at all, because it looks like the business has closed.
Other baseline numbers
- 17.9% have a public email address on file (the rest only show a phone number or address).
- 61.1% have a phone listed on Google Maps. The majority are landlines, not mobile.
- 4.2% have an Instagram handle attached to their Google Maps listing. The real figure of businesses with an active Instagram is higher; what Maps surfaces is a subset.
- Median sales-page fit score across the sample: 33 out of 100. There is broad opportunity, but half the businesses are not yet ready to buy a website, they are ready to be educated on why they need one.
03Gap by city
Bogotá, Medellín, and the Coffee Region together hold 56% of the businesses sampled. The gap is uneven: mid-sized cities and the coffee and llanero regions are far further behind than the big metros.
| City / area | n | No site | Dead site | No working presence |
|---|---|---|---|---|
| Yopal | 320 | 100.0% | 0.0% | |
| Riohacha | 97 | 100.0% | 0.0% | |
| Coffee Region | 797 | 82.8% | 5.6% | |
| Cúcuta | 88 | 77.3% | 8.0% | |
| Popayán | 88 | 73.9% | 8.0% | |
| Armenia | 144 | 63.9% | 16.7% | |
| Santa Marta | 119 | 58.8% | 16.0% | |
| Cali | 309 | 56.6% | 17.5% | |
| Manizales | 75 | 28.0% | 44.0% | |
| Barranquilla | 120 | 63.3% | 8.3% | |
| Bucaramanga | 126 | 61.1% | 10.3% | |
| Medellín area | 822 | 60.1% | 11.3% | |
| Bogotá | 1,390 | 54.1% | 15.8% | |
| Cartagena | 181 | 37.0% | 16.0% |
Yopal and Riohacha both come in at 100% no-site. Cartagena, the international tourist city, sits at the top of the table: 47% of its SMBs have a working site. Bogotá is not the best, it is just the biggest. Manizales is an interesting outlier: the lowest no-site rate of any metro (28%), but also the highest dead-site rate (44%). Manizales SMBs tried, then a lot of them abandoned what they built.
04Gap by sector
Sector explains more variance than city. Some sectors are almost entirely offline; others are partly so. Here is the sectoral picture, uncut:
| Sector | n | No site | Dead site |
|---|---|---|---|
| Corner store / supermarket | 151 | 100.0% | 0.0% |
| Pharmacy | 260 | 94.2% | 1.5% |
| Veterinary clinic | 122 | 89.3% | 2.5% |
| Hair salon / barber | 72 | 81.9% | 5.6% |
| Hardware store | 81 | 79.0% | 7.4% |
| Auto mechanic | 93 | 77.4% | 14.0% |
| Bakery | 80 | 75.0% | 6.2% |
| Fast food | 212 | 73.1% | 5.7% |
| Restaurant | 842 | 69.1% | 11.5% |
| Hostel | 172 | 52.3% | 26.7% |
| Hotel | 498 | 62.9% | 13.9% |
| Moto repair | 48 | 62.5% | 14.6% |
| Apparel | 89 | 65.2% | 3.4% |
| Clinic | 76 | 55.3% | 9.2% |
| Café | 170 | 54.1% | 16.5% |
The everyday sectors that touch the most Colombian lives sit furthest behind: pharmacies, vets, hair salons, hardware stores, mechanics, bakeries. Nearly 9 in 10 Colombian pharmacies have no website. Nearly 9 in 10 vets either. If your business is one of these sectors, having a simple site puts you ahead of 80% of your local competition on web presence, without doing anything fancy.
At the other end, sectors where customers book online (hostels, cafés, clinics, apparel) have better coverage. But watch the hostel row: 26.7% of hostels with a site have a dead one. That is the worst dead-rate of any sector we measured. A booking page that does not load is, in practice, a booking page that does not exist.
05State of the sites that do exist
Of the 1,871 businesses that do have a URL on file, here is the technology breakdown:
Two headlines here. First, 1 in 4 Colombian SMB websites is down: it counts as existing in Google but not as existing for the customer. Second, WordPress is the dominant platform at 27.5% of the stock. WordPress is not bad, but it has a known failure mode: when the owner stops paying for hosting, stops updating the theme, or forgets the site exists, the site breaks. The majority of the "dead sites" we found are ex-WordPress.
Technical note
CMS detection runs on HTML fingerprints (meta generator, asset paths, headers). It does not catch sites that have been hand-edited to hide their platform. The WordPress share is probably understated by 3 to 5 points.
06Technical audit of 60 live sites
To see how well-built the sites that do work actually are, we took a random sample of 60 URLs that responded 200 OK and downloaded their HTML directly. We looked for basic web-quality markers: HTTPS, a mobile viewport tag, a favicon, Open Graph tags for social previews, and server response time. Here is what we found:
Almost half the sites have no preview image for WhatsApp, Facebook, or Twitter. When someone shares the link, it shows up as a grey box.
One in three has no preview title either: the shared link goes out with no name and no description.
The server takes more than 3 seconds to send the first byte. On mobile, that loses half the traffic before content loads.
The basics that are in place:
- 98.3% have HTTPS. The era of "Not secure" in Chrome is over. Only one of 60 sites still ran on plain HTTP.
- 95.0% have a mobile viewport. Not because Colombian SMBs explicitly set it: WordPress, Wix, and Shopify add it by default. But the 5% that do not have it look broken on a phone.
- 90.0% have a favicon. The remaining 10% show the ghost icon in browser tabs.
The most actionable finding here is the Open Graph one. The number-one way Colombians share a business with a friend is to send a link over WhatsApp. If the link has no og:image and no og:title, WhatsApp shows a grey box with the raw URL. Rich link previews (a photo of the place, the business name, a one-line description) raise the click-through rate of a shared link by 2 to 3 times by industry-standard patterns. Nearly half of the audited sites do not have it configured. It is the cheapest possible win: three lines of HTML.
07The "Instagram only" pattern
A common question: don't SMBs make up for the lack of a website with social media? The data says: partly.
- 4.2% of the total (221 businesses) have an Instagram handle on file and no website. That is the figure Google knows; the real one (because owners often do not link their IG from Maps) is clearly higher, based on our qualitative sub-sample of Medellín clinics where roughly half of the no-site ones did have an active IG.
- Only a handful of businesses without a site have a formal WhatsApp number listed on Google Maps. Most use the phone field (a landline or unlabelled mobile) instead of the WhatsApp field on the Google profile.
- What we did not find: a single SMB with Instagram that also had a healthy website. Businesses that care about their site usually have IG too, but the "IG yes, site yes, both healthy" combination is rare in Maps.
The problem with "Instagram only" is discoverability. When someone Googles 24 hour pharmacy Belén Medellín, Instagram profiles do not show up in the results. Websites do. If the pharmacy has no site, that search goes to a pharmacy that does.
08What this all means
Three takeaways, in order of importance:
One. For the average Colombian SMB, existing on the web is already a competitive advantage. You do not need technical SEO, generative AI, or an optimised landing page with A/B testing. A site that loads, with a clear address, hours, and a WhatsApp button puts your business ahead of 76.7% of its peers.
Two. Sector matters more than city. A pharmacy in Bogotá is in worse shape relative to its sector (94.2% no-site nationally) than a hotel in Yopal is relative to its sector, even though Yopal is at 100% no-site. The opportunity to stand out is biggest where the category is furthest behind: pharmacies, vets, hair salons, hardware stores, mechanics, bakeries. There, a decent site is enough to win the local search.
Three. Old sites are a risk, not an asset. A site that loads slowly, has no preview image, and goes down periodically is actively pushing away customers who were already interested. 33.8% of existing sites are dead. If your business has a site you have not touched in three years, it is probably in that 33.8%.
Want to see what your business would look like with its own page?
We build a free mockup using your actual business content (your photos, your hours, your services) so you can see what it would look like. No charge to see it. If you like it, the page is live within 48 hours, from $400.000 COP.
PymeWebPro · Studio in Medellín · Colombian SMBs only, billed in COP
09Methodology and limitations
How the data was gathered
The 5,305 businesses analysed come from a primary extraction over the Google Places API, run by the PymeWebPro internal prospecting system between May and November 2025, deduplicated by business name and city, and consolidated in a Cloudflare D1 database. The "city" field uses the Google Places area names (for example "Medellín área" or "Eje Cafetero"), not the exact political division.
The "website" field comes directly from each business's Google Maps listing. If the owner did not add their URL to the Google profile, we count it as "no site" even if a site exists elsewhere. This is a real limitation: it understates true web presence, but it also measures something useful, which is the web presence a real customer can find through a normal Google search.
For businesses that did have a URL on file, we run a CMS detection on the downloaded HTML (meta generator, asset paths, headers) and classify the site as WordPress, Wix, Shopify, Drupal, Joomla, GoDaddy, Squarespace, Webnode, Jimdo, custom_or_other (loaded but with no recognisable fingerprint), site_dead (timeout or blank), site_unreachable (DNS or connection refused), 404_not_found (explicit 404), server_error (5xx), or a specific HTTP code (403, 410, 429, etc.).
The 60-site technical audit
For section 6 (HTTPS, viewport, Open Graph, performance) we drew a stratified random sample of 65 URLs from the database, picking only those not already flagged as dead, and forcing unique hostnames to avoid repeating chains. We fetched the home page with a research-bot User-Agent, a 5-second timeout, and a 200 KB cap on the HTML read, parallelised across 16 threads. Of 65 attempts, 60 returned HTTP 200; 5 failed on timeout, DNS error, or refusal. Reported results are over the 60 successful fetches.
This audit uses HTML download only, not full-page rendering in a real browser. So we report markers visible in the initial HTML (viewport, OG, favicon, HTTPS, time-to-first-byte) and we do not report metrics that require a real browser, like Largest Contentful Paint, Cumulative Layout Shift, or full DOM load time. That second layer requires Lighthouse or Playwright running from Colombian regions, and we will add it for the 2027 edition.
What this report does not say
- It does not measure traffic or sales. It measures web presence. A site can exist and get no visitors; another can be visited often. That is a different study.
- It does not measure conversion. We do not know how many visitors ended up messaging on WhatsApp or calling.
- It is not representative of all Colombian SMBs. The sample is drawn from Google Places, which overrepresents businesses that bothered to claim a Maps listing and underrepresents the fully off-the-grid informal market.
- Small samples are small samples. Sectors with fewer than 50 businesses in our data (electricians, plumbers, locksmiths) all came in at 100% no-site, but the sample is too thin to publish as a firm statistic. We left those out of the main tables.
Raw data and replication
This report is generated annually. The base extraction lives in the PymeWebPro internal database (pymewebpro-portal on Cloudflare D1) and the aggregates are versioned in the studio repository under _audits/2026-stats.json and _audits/2026-site-audits.json. If you work in market research and want access to the aggregated dataset by sector and city, write to hello@pymewebpro.com.
About the study
PymeWebPro is a Colombian web design studio focused exclusively on Colombian SMBs. We work on one question: how do you make it so that when a customer searches for your business, they find you and choose you? This report is part of our ongoing work to understand the base we work on: the real market, not the idealised one.