Desktop vs mobile Google results: what changes

Ask Google the same question on a laptop and on a phone, at the same moment and from the same country, and you may not get the same answer. The layout is different, which everyone expects. Less obviously, the results can be different too: which pages rank, in what order, and which extra features surround them.

If you work with search data, for rank tracking, SEO reporting or feeding results to an AI agent, this decides which version of Google you are measuring. This explainer covers what tends to differ, what we saw in our own side-by-side test, how to request each version with the Google Search API, and what it means for rank tracking.

Why the two versions differ

Google designs its results page for the screen it is shown on, and it can choose results with the device in mind. Some differences are well known and stable. Others change as Google experiments, so treat the following as general patterns, not rules.

• Layout. Desktop results have room for a second column, where Google often shows elements such as the knowledge panel beside the main results. On a phone everything sits in a single column, so those elements appear in the main flow and push other results down.
• How much is visible. A phone screen shows far less of the page at once. Being 5th can mean “visible without scrolling” on desktop and “several swipes down” on mobile, especially when an AI Overview, ads or a local pack come first.
• Titles and snippets. Google truncates titles and snippets to fit the space available, so the same page can look different on each device.
• Ranking itself. Google can rank pages differently by device. Google indexes most sites using the mobile version of their pages, but that doesn’t make the two results pages identical.
• SERP features. Features such as AI Overviews, People Also Ask, local packs and video results can appear on one device and not the other, or in a different order.

None of this has a fixed size. How much the two pages differ depends on the query, the country and the day.

What we saw in a side-by-side test

To ground this in data rather than generalities, we ran one query, project management software, in the US, once with device: "desktop" and once with device: "mobile". The two requests were made in October 2026, submitted within the same second. Here is what came back:

The headline is the organic results: not one domain appeared on both pages. The desktop page was led by Wrike, Teamwork.com and ICAgile, and included Zapier at 5 and Wikipedia at 7. The mobile page was led by ProofHub and ProProfs Project, with two Reddit threads and a Quora question in positions 3 to 5.

Not everything differed. Both pages had an AI Overview, and both cited exactly the same eight sources in the same order, though the overview text itself was not identical. The two devices disagreed completely on the organic results while agreeing on which pages to cite in the AI Overview.

The format of the results differed too:

// desktop, position 1
{
  "position": 1,
  "title": "Best project management tools and software: My 21 top ...",
  "link": "https://www.wrike.com/project-management-guide/faq/what-are-project-management-tools/",
  "displayed_link": "https://www.wrike.com › ... › FAQ › Tools",
  "snippet": "Aug 14, 2026 — What are the 11 best project management tools? ..."
}

// mobile, position 3
{
  "position": 3,
  "title": "Help a girl out – any project management software recommendations?",
  "link": "https://www.reddit.com/r/projectmanagers/comments/1q607zw/help_a_girl_out_any_project_management_software/",
  "displayed_link": null,
  "snippet": null
}

Two of the desktop titles were cut short with “…”, while none of the mobile titles were. Desktop displayed_link values were full breadcrumbs, while mobile ones were just the site URL, or null for the three forum results, which also had no snippet.

One query on one day shows that desktop and mobile results can be completely different. How much they differ for your keywords is easy to measure: the code below compares both devices for 2 credits per keyword.

How to request each version

The google-search endpoint takes a device parameter: "desktop" (the default) or "mobile". Everything else stays the same, and both cost 1 credit per request.

curl -X POST https://api.crawlerdata.com/v1/jobs \
  -H "Authorization: Bearer $CRAWLERDATA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "endpoint": "google-search",
    "params": { "query": "project management software", "country": "us", "device": "mobile" },
    "queue": "live"
  }'

The response has the same keys on both devices: organic_results, ads, ai_overview, people_also_ask, local_pack, top_stories, knowledge_panel and related_searches. Code that reads one will read the other. Write it to tolerate null in fields such as snippet, displayed_link and total_results, which in our sample were missing more often on mobile.

To compare the two for your own keywords, submit both versions at once on the fast queue and compare what comes back:

import os
import time
from urllib.parse import urlparse

import requests

API_URL = "https://api.crawlerdata.com/v1/jobs"
HEADERS = {"Authorization": f"Bearer {os.environ['CRAWLERDATA_API_KEY']}"}


def submit(query, device, country="us"):
    body = {
        "endpoint": "google-search",
        "params": {"query": query, "country": country, "device": device},
        "queue": "fast",
    }
    resp = requests.post(API_URL, json=body, headers=HEADERS, timeout=30)
    resp.raise_for_status()
    return resp.json()["job"]["id"]


def wait_for(job_id, every=10, max_wait=900):
    deadline = time.time() + max_wait
    while time.time() < deadline:
        data = requests.get(f"{API_URL}/{job_id}", headers=HEADERS, timeout=30).json()
        if data["job"]["status"] == "completed":
            return data["result"]
        if data["job"]["status"] == "failed":
            raise RuntimeError(data["job"]["error"])
        time.sleep(every)
    raise TimeoutError(job_id)


def both_devices(query, country="us"):
    jobs = {d: submit(query, d, country) for d in ("desktop", "mobile")}
    return {d: wait_for(job_id) for d, job_id in jobs.items()}


def host(url):
    return (urlparse(url).hostname or "").removeprefix("www.")


def compare(desktop, mobile):
    d_hosts = [host(r["link"]) for r in desktop["organic_results"]]
    m_hosts = [host(r["link"]) for r in mobile["organic_results"]]
    shared = sorted(set(d_hosts) & set(m_hosts))
    return {
        "organic_count": (len(d_hosts), len(m_hosts)),
        "shared_domains": shared,
        "rank_by_domain": {h: (d_hosts.index(h) + 1, m_hosts.index(h) + 1) for h in shared},
        "ai_overview": (desktop["ai_overview"] is not None, mobile["ai_overview"] is not None),
        "people_also_ask": (len(desktop["people_also_ask"]), len(mobile["people_also_ask"])),
        "related_searches": (len(desktop["related_searches"]), len(mobile["related_searches"])),
    }


serps = both_devices("project management software")
print(compare(serps["desktop"], serps["mobile"]))

For our test query this prints:

{'organic_count': (7, 10), 'shared_domains': [], 'rank_by_domain': {}, 'ai_overview': (True, True), 'people_also_ask': (4, 4), 'related_searches': (8, 4)}

Run it over a sample of your own keywords before deciding whether you need both devices. Two credits per keyword is a small price for knowing.

Why it matters for rank tracking

If you track rankings, the device is part of the measurement, not a detail.

Track them as separate series. A ranking record should be keyword + country + device + date. In our test, Zapier ranked 5th on desktop and was not among the 10 mobile organic results at all. Average the two and you get a number that describes neither page. Report desktop and mobile side by side instead.

Pick the device that matches your audience. Most analytics tools can split organic traffic by device. If most of your search visitors arrive on phones, mobile rankings are the ones to watch, and the other way round. Tracking only the default, desktop, is a common blind spot.

Position isn’t the same as visibility. Position 5 on a single-column mobile page under an AI Overview is a long way down the screen. If you record which features appeared above the organic results, such as ai_overview, ads and local_pack, you can tell a real ranking drop from a page that simply got more crowded.

Watch citations as well as rankings. In our sample, the AI Overview cited the same sources on both devices even though the organic results had nothing in common. A page can be absent from the mobile organic results and still be cited in the mobile AI Overview, as Zapier’s was. Recording whether your domain appears in ai_overview.sources captures that.

Mind the cost of doubling. Tracking both devices doubles the number of requests: one per keyword, country, device and page. At 1 credit each, 500 keywords on two devices in one country is 1,000 credits per daily check. The pricing page has a calculator for your own numbers.

Getting clean comparisons

• Compare at the same moment. Results change over time, so submit both devices together, as the script above does.
• Use the unpersonalised view as your baseline. Results show what Google returns for the query, country, language and device at request time, not what one logged-in user sees, so every check measures the same thing.
• Cover every market you report on. country takes one of 31 two-letter codes, and desktop and mobile cost the same 1 credit per request.

Next steps

To look at a single search before building anything, our free AI Overview checker runs without an account and shows the AI Overview, its sources and the top results for the country and device you pick.

To put this into practice, the SEO rank tracking page covers tracking positions by device and country. If you’re still choosing a data provider, our cheapest SERP API comparison covers the cost per 1,000 searches across the main options.