``Marketing doesn't know what they're doing.'' Before executives say this, it's important for marketers to create a system that can prove results with data. This chapter covers all the techniques for visualizing marketing contributions, from campaign management and multi-touch attribution to custom reporting and ROI dashboard design.
HubSpot campaign tools（マーケティング → キャンペーン) is a system that ``bundles, manages, and analyzes multiple measures (emails, SNS, advertisements, blogs, LPs) under one purpose.'' If you look at data for individual measures separately without using campaigns, you will never be able to see the overall ROI.
If the campaign names are different, it will not be possible to aggregate them during analysis.Example naming convention:2026Q2_新規リード_ウェビナー Let's unify the format in "Year/Quarter_Objective_Measure format" like this. HubSpot also has a "tag" feature for campaigns, so if you use tags to filter by purpose and channel, later analysis will be much easier.
In real B2B purchasing, it takes a long time for one prospective customer to receive an order. Average 6-10 touchpointsgo through. ``First ad seen,'' ``Blog article read,'' ``Webinar attended,'' ``Downloaded casebook,'' ``Last price page viewed'' -- all these things contribute to orders. Multi-touch attribution is a mechanism that allocates the contribution to each point of contact.
レポート → アトリビューションレポート or レポート → カスタムレポートビルダー → アトリビューション Create from (Professional or higher)No attribution model is the “one right answer.” Each is a tool that answers different questions.“Which channels are creating new awareness?” → First Touch、“What is driving the closing?” → Last touch、“I want to evaluate the overall contribution fairly” → Linear or U-shaped--When reporting to management, it is important to select and present a model according to the purpose.
The funnel report is a report that visualizes the number of people and conversion rate at each stage of "Visitor → Lead → MQL → SQL → Opportunity → Order".At what stage and how much have you left?By understanding this, you can see at a glance where there are bottlenecks in marketing and sales.
| Inter-stage conversion rate | Industry guideline (B2B SaaS) | Improvement actions if low |
|---|---|---|
| Visitor → Lead | 2〜5% | Optimize forms and CTAs / Review offers (downloads) / Improve LP (Chapter 3) |
| Lead → MQL | 20〜35% | Review of scoring rules / Strengthen nurturing sequence / Improve accuracy of form questions |
| MQL → SQL | 30〜50% | Review of consensus on MQL definition for marketing/sales/Improvement of sales follow-up speed |
| SQL → Opportunity | 40〜60% | Improving first approach messages/Improving accuracy in assigning to appropriate personnel |
| Business negotiation → Order | 25〜40% | Strengthening follow-up content after business negotiations/Analysis and countermeasures for reasons for loss of orders (Chapter 6) |
If you try to improve all stages of the funnel at the same time, the results will be dispersed and will not be effective. Conversion rate compared to industry standardsSelect one stage with the largest gap and focus on improving it.is the correct approach. In the example above, the conversion rate for "Lead → MQL" is 28%, which is within the standard range, but the conversion rate for "Visitor → Lead" is 5.2%, which leaves a lot of room for improvement.
For your company's unique analysis needs that cannot be covered by HubSpot's standard reports,custom report builder（レポート → カスタムレポートビルダー). It is a very flexible analysis tool that can aggregate across multiple objects such as contacts, companies, deals, and marketing activities.
| report name | data source | Main analysis contents | Reporting frequency |
|---|---|---|---|
| Lead acquisition cost by channel (CPL) | Contact × Advertising × Campaign | Compare lead acquisition unit cost by each channel (SEO, advertising, SNS, event) | monthly |
| MQL contribution by content | Contact × Content × Campaign | Which blog articles/LPs/downloads generate the most MQLs? | monthly |
| MQL by salesperson → Order conversion rate | Opportunity × Contact × Person in charge | How much each person in charge converts the MQL passed from marketing into orders? | Monthly/Quarterly |
| LTV by industry × channel | Business Negotiation × Company × Contact | Identify which industry orders have high LTV (lifetime customer value) and optimize advertising and content targeting | quarter |
| Webinar ROI | Campaign × Contact × Opportunity | Calculate how much revenue was generated against the cost of hosting a webinar | After the event |
Analysis does not have to be done once it is made, but has meaning only when it is used in management decisions.Dashboard for managementInstead of the operational indicators that marketers look at on a daily basis, only "top indicators that show business impact" should be included.
The first thing to show to management is Layer 1 (business results) is. Start with the following sentence: ``Order revenue from marketing was ¥18.4M this month, +23% compared to the previous month.'' The details of the measures in Layer 3 are used as evidence to explain why things happened, but if you go into details from the beginning, people will say, ``I don't know what you're trying to say.''
No matter how good a report you design,If the original data is dirty, the analysis results cannot be trusted.. “Garbage In, Garbage Out” is a golden rule of data analysis. Explain data quality management practices to improve analysis accuracy in HubSpot.
| problem | Symptoms/effects | How to deal with it |
|---|---|---|
| duplicate contact | The same person exists as multiple contacts and multiple nurturing emails are sent. Analysis numbers are inflated | 連絡先 → 重複を管理 Merge regularly from Enabling HubSpot's automatic duplicate detection feature |
| UTM parameters not set | The original source becomes "direct inflow" and you don't know which channel it came through. The contribution of advertising and SNS becomes invisible | Be sure to add UTM to all external links (advertisements, SNS, emails) (see naming rules in Chapter 1) |
| Lifecycle stage not updated | Nurturing emails are delivered to existing customers even though they are actually customers but remain as “leads” | Set up a workflow that automatically updates the negotiation stage to "Customer" in the life cycle when receiving an order (Chapter 6) |
| Contamination of employee/test contacts | Data from employee test submissions of company forms is mixed into analysis, distorting metrics. | Make Segment excluding employee domain (@company domain) the default filter for analysis view |
| Variation in property value notation | "Manufacturing", "Manufacturing", and "Mfg" are aggregated as separate values, and industry-specific analysis does not work. | Change the property to "Pulldown (fixed choices)" instead of free text. Modify existing data using bulk update tool |
HubSpot Breeze Intelligence(formerly Clearbit integration feature) has been enhanced from 2025 to 2026. This function automatically fills in missing information in contact and company properties (position, industry, number of employees, annual sales, etc.) from an external database.CRM records can be automatically enriched from the minimum information collected through forms (name, company name, email address)This has the effect of killing two birds with one stone by reducing the number of form fields and increasing conversion rates while maintaining data quality.
By linking email, LP, SNS, advertisements, and workflow to campaigns, ROI across measures can be calculated for the first time. Unification of naming rules is essential.
First touch, last touch, and U-shape are used depending on the purpose. Instead of deciding which model is correct, choose a model based on what you want to know.
Do not improve all stages at the same time. Concentrate resources on the stage where the conversion rate has the largest gap with the industry standard.
Order revenue, pipeline, and ROI are presented first as Layer 1. The policy details (Layer 3) will be used later as evidence.
Duplicate contacts, UTM not configured, and life cycle not updated are the three main causes of "dirty data". Incorporate weekly, monthly, and quarterly maintenance cycles.
Keep your CRM data rich while minimizing form fields. Balancing conversion rates and data quality will be the standard approach in 2026.