When Product Changes Become Customer Costs: An Ongoing SEMrush Case Study
As a digital strategist, I spend a significant amount of time evaluating software platforms—not just on features, but on how well they support real-world workflows over time.
Over the last year, I have found myself documenting an increasingly complex series of issues related to SEMrush’s API, project management tools, billing model, and customer support processes. Individually, none of these issues would have been particularly concerning. What has made the situation notable is the cumulative impact of repeated product changes, unanswered questions, and the amount of customer effort required to investigate and work around those changes.
Background
I have been a SEMrush customer for many years and have invested tens of thousands of dollars in the platform over that time. My agency uses SEMrush for SEO research, reporting, site audits, competitive analysis, and API-driven workflows.
Because SEMrush sits at the center of several business processes, seemingly small product changes can have meaningful downstream effects.
Issue 1: Changes to API Purchasing
Several years ago, SEMrush transitioned away from a bulk API credit model that allowed customers to purchase credits as needed. The replacement model relies on monthly allocations that expire if unused.
For organizations with highly variable API usage, this can create a disconnect between cost and actual consumption. In my case, I repeatedly communicated that my usage patterns were intermittent and that the previous model better matched my needs. Despite paying for API access over multiple years, many credits went unused while later projects encountered usage limits.
The core question I raised was simple:
How should long-term customers evaluate the value of a usage-based product when unused capacity expires and cannot be applied to future needs?
Issue 2: Removal of Bulk Project Deletion
One of my workflows relied heavily on creating and removing large numbers of projects through SEMrush.
At some point, the bulk deletion feature was removed.
The replacement workflow required either:
- Deleting projects individually
- Building a custom API process
- Sending project IDs to support for manual deletion
In practice, this resulted in repeated support tickets containing long lists of project IDs simply so work could continue.
The cost of the removed feature was not measured in dollars. It was measured in friction.
Issue 3: Questions About API Billing Behavior
While investigating API usage, I noticed several situations where credit consumption appeared unclear.
Among the questions I raised:
- Why some snapshot requests appeared to consume credits twice
- Whether failed project creation attempts consumed credits
- Whether failed audit launch attempts consumed credits
- Whether audit concurrency limits were clearly documented
These questions were accompanied by screenshots, examples, and follow-up requests.
The challenge was not discovering potential issues. The challenge was obtaining definitive answers.
The Larger Problem
The most interesting part of this experience is not any individual bug, billing question, or missing feature.
It is the amount of effort required to understand what was happening.
Over the course of the year I found myself:
- Investigating platform behavior
- Reviewing logs and usage reports
- Building test cases
- Creating screenshots and videos
- Writing lengthy explanations
- Following up repeatedly on unresolved questions
For a software platform, this represents a hidden cost that rarely appears on an invoice.
Lessons for SaaS Companies
This experience has reinforced a few principles that apply well beyond SEMrush:
- Product changes create customer costs, even when pricing remains unchanged.
- Removing workflow features often has greater impact than adding new ones.
- Documentation and transparency are often more valuable than compensation.
- Escalation paths matter when issues become complex.
- The cost of investigating a problem can exceed the cost of resolving it.
Software products evolve. Bugs happen. Features come and go.
What customers tend to remember is whether their time was respected throughout the process.
Status
At the time of writing, several questions remain unresolved and discussions with SEMrush are ongoing.
I am documenting the experience not because it is unique, but because it highlights a challenge that many software companies face as they grow: balancing product evolution with the needs of long-term customers whose businesses depend on stability, predictability, and clear communication.