ResearchMatch.site did not have a content problem in the usual sense. There were already useful pages on the site, including guides answering the kinds of questions students ask when they are trying to find professors, approach them about research and work out how to get their first research opportunity.
The bigger problem was that Google was not making enough use of them.
Some pages were not being indexed at all. Others were indexed but had yet to build meaningful visibility, while parts of the blog needed tightening so that the main answer was easier to understand and the articles were more clearly aligned with what students were searching for.
Our work therefore started with the existing site rather than a large new publishing plan. Over the following months, we worked through previously unindexed pages, tightened the blog library, submitted ResearchMatch to relevant directories and promoted useful content through Reddit. The resulting growth was most obvious in search visibility: ResearchMatch went from roughly 111 Google impressions a day during the first 30 days of the period to more than 600 a day during the final 30 days.
That is the part of the case study we think is most interesting, because it came largely from making better use of work that already existed.
About ResearchMatch.site
ResearchMatch helps students find professors whose research overlaps with their interests and gives them information they can use before reaching out. The site also publishes an extensive library covering undergraduate research, cold emailing professors, finding research positions and approaching potential mentors.
From an SEO perspective, that creates a fairly natural opportunity. Students ask an enormous number of specific questions while trying to get involved in research, and many of those questions are closely related to what ResearchMatch itself helps them do.
The challenge was getting more of that useful content surfaced in search.
The starting problem: good pages that Google was barely using
Our first concern was indexation. There is little point commissioning another twenty articles if useful pages that have already been written are not appearing in Google’s index. We therefore spent time identifying pages that had previously not been indexed and working through them so that Google could actually discover and surface them.
At the same time, we reviewed the existing blog. Some articles had useful information but could be more focused. We tightened the writing around the search question, reduced unnecessary padding and made the main purpose of each page clearer. The goal was not to make every article longer or stuff more keywords into it. In several cases, the better edit was simply a more direct article.
This was particularly important for ResearchMatch because the strongest topics are fairly specific. Someone searching for the best time to email a professor is looking for a different answer from someone trying to work out how to follow up after receiving no reply. The closer each article stays to that actual problem, the more useful the overall library becomes.
We also worked on discovery outside the site
The second part of the work happened away from ResearchMatch itself.
We submitted the site to relevant directories so that the product had a wider footprint beyond its own domain. We were selective rather than treating directory submissions as a numbers exercise; the objective was to create legitimate places where students, researchers and people looking for academic tools could discover the product.
We also used Reddit. That did not mean dropping a ResearchMatch link into as many subreddits as possible. Reddit works badly when a business treats it as a free advertising channel, particularly in communities where people are already sensitive to self-promotion.
Instead, we used relevant discussions as another way to get the site and its resources in front of people already talking about finding professors, research experience and cold outreach.
We would not attribute Google ranking improvements directly to a directory listing or a Reddit post. There are too many moving parts in organic search to make that claim responsibly. What those activities did give ResearchMatch was broader distribution and a more diverse footprint than relying on Google alone.
What happened to Google visibility
The clearest change in the Search Console data was impressions.
Across the three-month export from 12 July to 11 October 2026, ResearchMatch recorded:
| Metric | Result |
|---|---|
| Google Search clicks | 1,339 |
| Google Search impressions | 40,679 |
| Overall CTR | 3.29% |
| Impression-weighted average position | ~6.8 |
The more revealing comparison, however, is between the beginning and end of the period.
First 30 days vs final 30 days
| Metric | First 30 days | Final 30 days | Change |
|---|---|---|---|
| Impressions | 3,319 | 18,226 | +449% |
| Average impressions/day | 111 | 608 | +449% |
| Clicks | 400 | 426 | +6.5% |
| Average clicks/day | 13.3 | 14.2 | +6.5% |
| Approx. average position | 8.4 | 6.0 | Improved |
That distinction matters. If we only reported clicks, the result would look fairly modest. Search visibility tells a different story.
ResearchMatch was appearing in Google far more often than it had been at the start of the period, and its average position had improved at the same time. The site was beginning to compete for a much broader set of searches rather than depending predominantly on people who already knew the ResearchMatch name.
Clicks had not yet grown at the same rate.
We think that is worth showing rather than hiding. SEO improvements often happen in stages: pages become indexed, then begin earning impressions, rankings move, and only then does a larger proportion of that visibility turn into clicks and ultimately customers.
A case study does not become more useful by pretending that every metric moved by 400%.
Several blog articles started doing meaningful work
The page-level data also shows why improving the existing library mattered. One of the strongest examples is the guide on the best time to email professors.
Over the three-month period, that single article generated:
15,710 impressions
246 clicks
Average position: 4.84
That means the article was no longer simply another page sitting in the ResearchMatch blog. It was regularly being surfaced to students searching around a problem that sits very close to the product itself.
Other articles were beginning to build useful visibility too.
The guide on how to follow up with a professor recorded 5,356 impressions and 46 clicks, with an average position of 7.51.
The article on how to get research experience as an undergraduate generated 2,731 impressions and 32 clicks, with an average position of 9.25.
The guide asking whether professors respond to cold emails recorded 1,376 impressions and 23 clicks, with an average position of 6.74.
These are useful searches for ResearchMatch because they sit within the same journey as the product. Someone learning how to identify a professor, approach them and follow up is much closer to the problem ResearchMatch solves than someone arriving through a loosely related high-volume keyword.
That is the kind of traffic we would rather build.
Why did CTR fall as impressions grew?
There is one metric in the data that could look concerning without context: the first 30 days had a considerably higher click-through rate than the final 30.
We do not see that as evidence that the work failed. ResearchMatch’s early search traffic contained a strong branded component, including searches for the ResearchMatch name itself. Branded searches naturally tend to produce high CTRs because the person already knows which site they want.
As Google started showing ResearchMatch for a much larger number of informational searches, impressions grew much faster than clicks.
That is a fairly different problem to having no visibility at all.
The next stage of the work is therefore less about getting pages to appear and more about improving how much of that growing visibility turns into traffic. Pages already receiving thousands of impressions give us much better information about where title changes, stronger search snippets, content updates or movement from positions six to three could have meaningful impact.
There is now something to optimise.
What we would do next
The Search Console data gives ResearchMatch a clearer roadmap than it had at the beginning.
Rather than immediately producing a large batch of new articles, we would continue working on pages that are already showing promise.
There are articles sitting within striking distance of stronger page-one positions. There are pages receiving large numbers of impressions but relatively modest CTRs. There are also search queries appearing around the existing content that can tell us where an article needs a better answer or where a genuinely new page is warranted.
The article on following up with professors, for example, is already averaging around position seven. The undergraduate research guide is around position nine. Those are much more interesting SEO opportunities than starting with a completely unrelated topic at position nowhere.
We would also continue improving the relationship between informational content and the ResearchMatch product itself. Search traffic is useful, but the commercial question is whether students who discover one of these guides understand that there is a tool available to help them find professors and research opportunities.
That connection needs to feel natural rather than turning every blog article into an advertisement.
What we learned from the ResearchMatch SEO work
The biggest lesson was not that ResearchMatch needed more content.
It needed more of its existing content to count.
Getting previously neglected pages indexed gave Google more material to work with. Tightening existing articles made some of those pages stronger search results. Directory submissions and Reddit gave the product additional routes for discovery outside its own website.
There is a tendency in SEO to jump immediately to producing more: more articles, more landing pages, more backlinks and, increasingly, more AI-generated content. Sometimes the better starting point is looking carefully at what has already been built and asking why Google is not using it.
For ResearchMatch, that produced a 449% increase in average daily search impressions between the first and final 30-day periods, while average search position improved from roughly 8.4 to 6.0.
The site now has a much larger search footprint than it started with.
The next job is turning more of that footprint into clicks, product discovery and users.
How we approach SEO at The Tofu Media Group
This is also how we prefer to approach SEO more broadly.
We do not begin by assuming a client needs another monthly content package.
We look at what Google has indexed, what it has ignored, which existing pages are already earning impressions, where content can be tightened, how the site is being discovered elsewhere and which changes are most likely to create momentum before adding more work.
ResearchMatch is a useful example because the growth did not come from publishing hundreds of new pages. Much of it came from making existing pages easier for Google to find and more deserving of the visibility they received.
If your website has useful content but Google seems to be ignoring large parts of it, you can submit the site through our Free Growth Review. We will look at what stands out, where we would investigate first and what we would prioritise before spending more on content or SEO.