Outbound Is Saturated: How Decision-Makers in IT & Software Find Vendors Today

Short answer: Software decision-makers increasingly start their vendor search in AI chatbots and on review sites, not with a reply to your cold email. According to G2, 50% of B2B software buyers begin their research in a chatbot instead of Google, and review sites (38%) and AI chats (37%) shape shortlists the most. If you want inquiries in IT and software, you therefore need pages that answer buyer questions directly, and proof that a buying group with an average of 13 internal participants can check.

The key points in brief

• The most reliable numbers on AI research come from IT and software, and the market is so large that many vendors compete for the same decision-makers.
• A software decision rarely has a single addressee. Tech, business units, security and purchasing ask different questions of different pages.
• Review portals, comparison pages and integration pages often decide the shortlist before your sales team hears about it.
• A simple example calculation shows: a landing page conversion through ads quickly costs several hundred US dollars in the software space, and a cold email reply is not yet a qualified demo.
• Outbound remains useful as small, targeted outreach. As the only channel for new customers it is risky in this sector, because many vendors approach the same roles.

This guide goes deeper on the pillar topic Outbound Is Saturated: Why Decision-Makers Find Vendors Through AI Searches. You'll find the general rates for cold email and LinkedIn there. Here we cover what is different in the software industry: the buying group, the shortlist sources and the page types a SaaS vendor needs.

Why IT & Software in Particular Is Saturated

A channel becomes saturated when many vendors approach the same people at the same time. There are several indications of this in the software industry, even though a direct measurement of message volume is missing.

The market is big and growing fast. Bitkom forecasts software revenue of €58.1 billion in Germany for 2026, up 9.9%. Cloud software accounts for €42.5 billion of that (up 21.9%), AI platforms for €3.1 billion (up 75.8%) (Bitkom, July 7, 2026, based on IDC data and Bitkom studies). More revenue usually also means more vendors and more sales teams approaching the same managing directors, IT leads and business unit heads. That is a conclusion drawn from market size, not a measured figure for message volume.

The paid B2B channel is shaped by SaaS. In an analysis of LinkedIn ad accounts, 97% of customers were B2B SaaS companies, with about $700,000 in ad spend between May 2025 and May 2026 (seo-day.de, based on a Search Engine Land article). The sample comes from a single agency and is small. It still shows who is bidding on the same job titles in this channel. The average cost per click in the analysis was $11.12 (LinkedIn Ads) and $12.48 for Google search ads without a brand term.

Volume lowers the rate. The Hunter data applies across industries, but the pattern is especially relevant for software vendors, because software sales teams often work with sequencing tools (an assumption, not a measured value): campaigns with up to 50 recipients reach a 5.8% reply rate, campaigns with 1,000 or more recipients only 2.1% (Hunter, State of Cold Email 2025, survey of 217 decision-makers). On LinkedIn, the reply rate to connection requests fell from 3.5% to 2.2% within twelve months (Expandi, May 2025 to April 2026). Both vendors sell outreach software themselves, so the figures are tool-maker benchmarks.

Buyers push back. According to Gartner, 73% of B2B buyers avoid vendors that send them irrelevant outreach messages (Gartner, June 2025). For a software vendor that lives on referrals and reviews, that is a double risk: bad outbound doesn't just save nothing, it can damage your reputation.

A reply rate of 2 to 6% doesn't sound bad with a good list. The problem: a reply is not a demo, and a demo is not a closed deal. The following calculation shows this.

Roughly What an Inquiry Costs in Ads and Cold Email

The following is an example calculation with assumptions, not an industry average. It combines values from different sources and is meant to show how quickly the orders of magnitude shift. Plug in your own values, ideally in the cost-per-lead calculator.

Channel	Assumed cost per click or reach	Assumed conversion	Calculated per conversion	Sources
Google Ads (search ad, no brand term)	$12.48 per click	5.1% (median of SaaS landing pages with Google search ads)	about $245	seo-day.de (CPC), Unbounce (conversion)
LinkedIn Ads	$11.12 per click	3.8% (median of SaaS landing pages overall)	about $293	seo-day.de (CPC), Unbounce (conversion)
Cold email, mass sending	1,000 emails	2.1% reply rate	21 replies	Hunter
Cold email, small campaigns	20 campaigns of 50 emails (also 1,000 emails)	5.8% reply rate	about 58 replies	Hunter

How to read the table:

• A conversion is not the same as a qualified lead. Unbounce measures landing page conversions across all goals, including whitepaper downloads. The SaaS median is 3.8%, all industries together 6.6%. The value varies widely by traffic source: email 16.9%, Google search ads 5.1%, Facebook 3.5%, display 0.3% (Unbounce Conversion Benchmark Report, probably the 2024 edition, the page doesn't state the year explicitly). If only every third lead truly fits your target customer, which is just an assumption for illustration, the price per qualified lead roughly triples to $700 to $900.
• A reply is not a demo. Hunter counts every reply, including "not interested." The study doesn't report how many replies lead to a meeting.
• The currency and the sources don't match cleanly. The click prices come from a mostly US-heavy SaaS sample, the conversion rates from a different dataset. There is no reliable series of click prices for the DACH region.

The result is not a number to quote but a method: calculate per channel with your cost per click, your conversion and your qualification rate. Those who do this often notice that with every additional euro, paid channels and outbound deliver only a fraction of the effect the first euro had.

How Software Buyers Research Today

Now to the other side. The most reliable numbers on AI research come from the software sector, because vendors like G2 survey their buyers directly.

The chatbot increasingly replaces the first Google click. In August 2025, G2 surveyed more than 1,000 B2B software buyers. The result: 87% say AI chatbots are changing their research, 50% start the buying process in a chatbot instead of Google, and 47% name ChatGPT as their preferred model. According to G2, the 50% figure is a 71% increase compared with a survey about four months earlier (G2, learn.g2.com). The study names neither region nor company size, and G2 itself runs a review portal. Treat it as a directional statement.

A later survey confirms the direction: in a summary by The Starr Conspiracy that G2 cites, 51% of software buyers more often start with an AI chatbot than with Google (previous year 29%, G2, April 2026). For shortlists, the same summary names review sites at 38% and AI chatbots at 37% as the most important sources (G2, July 2026). The sources differ in methodology and question wording, so the values are not directly comparable. For you that means: the shortlist comes from two sources you can only influence indirectly, third-party reviews and AI answers that draw on many sources.

Gartner and Forrester show the same pattern from another angle. 45% of buyers used generative AI in their last purchase, mainly for vendor and product research, 70% prefer a fully digital self-service experience, and 69% still want to validate AI results with a vendor or sales rep (Gartner, May 2026, survey August to September 2025). So your window is validation: the candidate is already in the AI answer, and your sales team or your website has to confirm it.

The decision rests on many heads. Forrester cites 13 internal participants and 9 external influencers for a typical purchase decision (State of Business Buying 2026). For purchases with GenAI features, the buying group grows to 14 instead of 7 members according to Forrester (Forrester, January 2026, cited by The Starr Conspiracy). That is a hint that AI products need more alignment because more roles have a say: data protection, security, business unit, IT.

And they decide early. According to 6sense, 94% of buying groups had ranked their preferred vendors before the first vendor contact, and the favorite wins in about 80% of cases (6sense B2B Buyer Experience Report 2025, over 4,000 buyers, the press release gives 77% elsewhere). These values have not been tested for German conditions.

The Buying Group: Who Asks the AI What?

The pillar page explains that decision-makers research in advance. In the software industry, the next step is worth taking: each role in the buying group asks a different question, and each question needs a different page. The table is a derivation from the study findings on group size and from typical buying processes, not a measurement.

Role	Typical question to search or AI	Page that has to answer	Proof that convinces
Management, business unit lead	"Which tools for [process] exist for companies like ours?"	Use case page per industry and process	Customer example with starting point and result
IT lead, developers	"Does [tool] have an API / SSO / interface to [system]?"	Integration and documentation pages	Public docs, changelog, status
Data protection, security	"Is [tool] GDPR-compliant (DSGVO), where is the data stored?"	Security page, data processing agreement, subprocessors	Certificates, server location, contracts to download
Purchasing, finance	"Roughly what does [tool] cost, what are the contract terms?"	Pricing logic, package comparison, contract terms	Transparent price structure or clear logic, notice periods
Users, key users	"How long does onboarding take, what does day-to-day look like?"	Onboarding page, product tour	Screenshots, videos, rollout plan

If one of these rows is empty for you, that is your lever. The AI answers the question anyway, just with a competitor's page or a review platform.

Which Pages a SaaS Vendor Really Needs

Most software websites have a homepage, three feature descriptions, a pricing page and a contact form. That is enough for visitors who are already convinced. For the research phase you need more, separated by question.

Feature pages as an answer, not a brochure. A feature page answers in its first sentences: what can it do, for whom, with what result? Then come the screenshot, use case and limits. Naming limits sounds risky, but it increases credibility and gives AI systems an honest, citable statement.

Integration pages, one per system. "Does [tool] connect with [CRM / ERP / ticketing system]?" is one of the most common buyer questions. A dedicated page per integration answers it directly and can show up as a source for questions about interfaces. We show how integrations are built into websites in the guide on CRM integration.

Comparison and alternatives pages. Buyers ask: "[Vendor A] or [Vendor B]?" and "Alternatives to [Vendor A]." If you don't write these pages yourself, you leave the assessment to review portals. Comparing fairly means naming the competitor's strengths, showing your difference factually, and saying for whom each solution fits. Buyers spot aggressive comparisons without proof.

Security and data protection page. It is not a box-ticking exercise but a sales page for security and purchasing. Server location, encryption, data processing agreement, certificates (if available), contact person. The page must be reachable without a login.

Pricing logic. The Contentful study (900+ decision-makers in manufacturing, industry and tech in EMEA) names lack of price transparency as the most common frustration (39%). That applies to the total sample, not specifically to software. The direction is still plausible: whoever says nothing about the price structure drops out with buyers who need an assessment without a conversation. You don't have to publish a price list. Explain what the price is based on (users, modules, volume) and what a typical rollout includes. How to handle prices on a website is covered in the post Show Prices on Your Website.

References with context. A logo bar is not a reference. A reference names the industry, the starting point, the rollout time and what changed. Only state figures the customer has approved. More on trust elements in the post Trust Elements on the Website.

What the structure of a B2B site can look like overall is described in the post Requirements for a B2B Website. For explaining complex software in the hero area, there is the guide Explaining Software in the Hero.

Review Portals and Third-Party Sources: Why Your Own Website Isn't Enough

Because review sites lead the shortlist with 38% according to G2, it is not only your website that counts, but also your profile on the relevant portals. That is what sets the software industry apart from almost every other.

What you can do there:

1. Keep profiles complete. Category, description, prices (if possible), integrations, screenshots, current logos.
2. Collect reviews systematically. After a successful onboarding, not after a support case. A short email with a direct link is enough.
3. Respond to reviews. Negative ones too, factually and with an offer to fix things. Future buyers read them.
4. Create consistency. Product name, core message and target group should match on your website and the portals, so assistants don't assemble a contradictory picture.

There are no reliable studies for Germany on whether AI systems cite portals more often than vendor websites. What makes sense in any case: make sure every source an assistant can draw on says the same thing.

From Interest to Inquiry: The Demo Path

According to Gartner, 70% of buyers prefer a fully digital self-service purchase. For SaaS that means: make the first step small before you ask for a conversation.

Typical entry points in order of increasing commitment:

• Interactive product tour or recorded demo without a form
• Trial access or sandbox, if the product allows it
• Price or effort calculator
• Whitepaper, checklist or template (lead magnet with a qualification question)
• Demo booking with a calendar instead of a contact form

The source strongly influences conversion: email traffic converts best on SaaS landing pages at 16.9%, display worst at 0.3% (Unbounce). For visitors from AI searches there is only a single case: Ahrefs reports that visitors from AI searches converted at about 23 times the rate (0.5% of visitors, 12.1% of sign-ups). The author also warns against generalizing (Ahrefs, visitors from AI search and conversion).

Practical tips for the contact path are in the post Contact Page in B2B and in the guide Optimize Your Inquiry Form.

Legal Framework for Cold Outreach in Germany

One specific point for the German market that doesn't appear in international benchmarks: § 7 UWG (German Unfair Competition Act) regulates advertising by phone and email. Phone advertising to consumers requires prior express consent, while for other market participants presumed consent is enough. Advertising by email is always considered an unreasonable nuisance without the recipient's prior express consent. An exception applies to existing customers whose address was collected in connection with a sale, for the company's own similar products and with a clear notice of the right to object (§ 7 paras. 2 and 3 UWG).

This is not legal advice. For you it follows: Hunter and Expandi rates come from markets with partly different law. A German SaaS vendor that cold-emails is on thinner legal ice than the benchmarks suggest. That is one more argument for visibility where the buyer contacts you. If in doubt, have the practice reviewed legally.

How to Test How AI Assistants Answer for Your Software Category

Run the test before you build pages. Ask the questions a buyer would ask and note who appears.

Example prompts for a B2B software product:

1. "Which software for [process] suits a company with [size] in [industry]?"
2. "Alternatives to [well-known competitor] for the German market"
3. "[Your product] or [competitor]: differences and who is each one for?"
4. "Does [product] have an interface to [system] and is it GDPR-compliant (DSGVO)?"
5. "What should I look for when choosing [software category]?"

Repeat the questions in ChatGPT, Perplexity and Google AI Mode, each time with an empty chat history. Record: Are you named? At what position? Which sources are shown? That is your baseline. How to measure this on an ongoing basis is explained in the post Measure AI Visibility, and how visibility comes about in the post Show Up in ChatGPT.

What You Can Do in 90 Days as a Software Vendor

1. Weeks 1 to 2: Collect questions. Ask sales, support and customer success: Which questions come up before every demo? Which objections keep repeating? That is your editorial plan. Add to it with B2B keyword research.
2. Weeks 3 to 6: Close gaps in page types. Start with the three pages that appear most often in buyer questions: usually integration, comparison and security.
3. Weeks 5 to 9: Align profiles and proof. Review portals, references, certificates and website say the same thing.
4. In parallel: Shrink outbound. Fewer recipients, a clean reason for reaching out (tech stack, job posting, event), a link to the matching expert page instead of the homepage. The Hunter numbers (5.8% versus 2.1%) support this route.

After 90 days, check: Does your name show up in the test prompts? Do inquiries come in with a reference to AI or portals? Does the form field "How did you find us?" match your expectations?

What You Can Handle Yourself Without an Agency

Not everything needs outside help. Test prompts, review profiles, collecting buyer questions and the content of the integration and security pages are things a good marketing team can handle in-house, because the expertise is with you. External support pays off mainly for structure and technology: information architecture for many page types, a clean template system, fast page builds, structured data, consistent brand management across website, docs and portals. An outside view that finds blind spots often comes on top.

When Outbound Still Stays for Software Vendors

An honest assessment belongs here:

• New category. If nobody searches for your solution, no AI can name you. Then you have to create the demand yourself, with content, events, partners and targeted outreach.
• Account-based sales with a small list. With 50 to 200 clearly defined target accounts, real preparation beats any sequence. The principle only works with research per account.
• Warm outbound. Whoever can refer to an event (webinar attendance, trade fair contact, referral) is no longer writing a cold contact.
• Short-term pipeline needs. Ads and outreach deliver faster than content. That is legitimate as long as you build substance in parallel.

Conversely: the data on AI recommendations and B2B orders is young and comes mostly from the US and from vendors that sell something themselves. The direction is well supported, the order of magnitude for your market is open.

How We Implement This

Webwavers is a premium agency for branding, web design and web development with locations in Günzburg and Munich, and one of seven Webflow Premium Partners in Germany. For software companies, we build websites with a clear information architecture in which feature, integration and comparison pages can each be found separately. One example from the sector is the rebranding of ScaleITs, a company for IT optimization and automation, where the existing brand was developed further without a restart and now presents itself consistently across all touchpoints. We advise neutrally on which platform fits (Webflow, WordPress, TYPO3 or headless). You'll find more on the topic on the page GEO and AI Visibility, and the general project process on the page Process.

Frequently Asked Questions

Does cold outreach still work for SaaS companies at all?

Yes, but only on a small, well-prepared scale. According to Hunter, campaigns with up to 50 recipients reach a 5.8% reply rate, campaigns with 1,000 or more recipients only 2.1%. On LinkedIn, the reply rate to connection requests fell from 3.5% to 2.2% within twelve months (Expandi). In Germany there is also the legal hurdle for email advertising without prior consent (§ 7 UWG).

How do software buyers research vendors today?

Increasingly through AI chatbots and review portals. According to G2, 50% of software buyers start their research in a chatbot instead of Google, and review sites (38%) and AI chatbots (37%) shape shortlists the most (G2, 2025 and July 2026). According to Gartner, 69% then want to validate the AI results with a vendor.

How many people have a say in a software purchase?

Forrester cites typically 13 internal participants and 9 external influencers. For purchases with generative AI features, the buying group grew from 7 to 14 members according to Forrester (January 2026). So plan pages for every role: tech, business unit, security, purchasing.

Which pages does my SaaS website need so AI assistants name it?

A dedicated page for every recurring buyer question: use cases, integrations, comparisons and alternatives, security and data protection, pricing logic, references. Google states that for AI features, normal SEO fundamentals essentially apply: indexable pages, important content as text, good internal linking (Google Search Central).

Should I stop my ads if I want to become visible in AI answers?

No. Visibility takes months, while ads deliver immediately. Keep ads for search terms with clear purchase intent and build content in parallel. Use the cost-per-lead calculator to check when the build-up pays off with your own values.

How do I measure whether ChatGPT and Perplexity name me as a vendor?

With fixed test prompts that you repeat regularly, with referral data in your analytics reporting, and with the question in the demo form about how someone found you. The method is described in the post Measure AI Visibility.

Does this also apply to IT service providers and system integrators, not just software?

The pattern applies there too, but the evidence is thinner, because the G2 numbers concern software buyers. For service providers, references, certifications, industry knowledge and contact persons count on top. The pages are similar: service pages per topic, industry pages, case studies.

Conclusion

In the software space, outbound is especially prone to saturation, because the market is large and many vendors approach the same roles. At the same time, buyers increasingly start in AI chat and read reviews before they contact a vendor. A buying group with many participants needs a page for each role that answers its question. Whoever is missing there has to compete against the favorites. The first step is a test with five buyer questions in three AI assistants, which you can run today.

Support With AI Visibility for Software Companies

If you want to know how your website performs on vendor questions in AI assistants, get in touch via the contact page. The industry page IT, Software & SaaS gives you an overview of the topic.

Sources

1. G2: AI search surging for B2B buyers (August 2025)
2. The Starr Conspiracy: B2B Buying Committee Benchmarks 2025 (G2, Forrester, Gartner summary)
3. Gartner: 69% of B2B buyers validate AI insights with sales (05/20/2026)
4. Gartner: 61% of B2B buyers prefer a rep-free buying experience (06/25/2025)
5. Forrester: The State of Business Buying 2026 (01/21/2026)
6. 6sense 2025 B2B Buyer Experience Report, press release (11/12/2025)
7. Bitkom: Digitalbranche weiterhin stabil (07/07/2026) (Digital industry remains stable)
8. Hunter.io: The State of Cold Email 2025
9. Expandi: LinkedIn Outreach Benchmarks 2026
10. seo-day.de: LinkedIn Ads CPC Benchmarks (report based on Search Engine Land)
11. Unbounce: SaaS Conversion Benchmark Report
12. Contentful: 2025 B2B Buyer Benchmark Report
13. Ahrefs: Visitors from AI search and conversion
14. Gesetze im Internet: § 7 UWG, Unzumutbare Belästigungen (Unreasonable nuisances)
15. Google Search Central: AI features and your website
16. Webwavers: ScaleITs project